{"name":"Policy Window — thematic research synthesis","description":"Per-theme cross-article dossiers: cross-jurisdiction instrument stances (governs/implicit/conflicts + provision excerpts) and evidence convergence (sources cited across multiple articles, a scientometric consensus signal computed from inline prose citations). Deterministic.","url":"https://policywindow.org/wiki/synthesis.json","license":"https://creativecommons.org/licenses/by/4.0/","citationLicense":"CC0 1.0 (citation data, per I4OC)","attribution":"Source: Policy Window AI Governance Catalog (https://policywindow.org/wiki).","catalogContentHash":{"algorithm":"sha256","hash":"27f88c1d52b0836d71db03c6da9b36adf9db9e383e5bfb2fadb0081f2ff6a4ff","cellCount":525,"recipe":"SHA-256 (hex) of the published coverage cells rendered as one line each, `<instrument>:<topic>\\t<type>\\t<citation>\\t<confidence>` (confidence empty if unset), the lines sorted ascending and joined with '\\n'. Re-derive from /wiki/catalog/json: for each entry in `.coverage` emit the same line, sort, hash."},"summary":{"themes":23,"totalStances":335,"themesWithConvergence":23,"totalConvergedSources":669,"maxConvergence":44},"dossiers":[{"topicCode":"agentic_systems_governance","label":"Agentic AI Governance","url":"https://policywindow.org/wiki/synthesis/agentic-systems-governance","themeArticle":"https://policywindow.org/wiki/agentic-systems-governance","empiricalConsensus":"emerging","contestedQuestion":"Should governance attach to the AGENT (multi-step actions, tool use, recursion) or to the model that powers it? Capability-tier vs action-tier frames are unresolved across jurisdictions.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"implicit","citation":"Arts. 26-29 deployer obligations apply to agent operators; Arts. 51-55 GPAI obligations capture the underlying model","excerpt":"Deployers of high-risk AI systems shall take appropriate technical and organisational measures to ensure they use such systems in accordance with the instructions for use accompanying the systems…","isParaphrase":false},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"implicit","citation":"Arts. 4, 8 (service-provision scope) — agent-like generative services fall within registration + safety-assessment obligations","excerpt":null,"isParaphrase":false},{"shortCode":"G7-HIROSHIMA","name":"G7 Hiroshima AI Process Code of Conduct","slug":"g7-hiroshima-code","verdict":"implicit","citation":"Code §1 'advanced AI systems' + §3 risk-identification cover agentic behaviour through capability frame","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"implicit","citation":"General-AI scope (Art. 3) covers agent systems; no agent-specific provision","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF","name":"NIST AI Risk Management Framework","slug":"nist-ai-rmf","verdict":"implicit","citation":"Map / Manage functions apply to autonomous systems; no agent-specific profile yet","excerpt":"Mechanisms are in place and applied, and responsibilities are assigned and understood, to supersede, disengage, or deactivate AI systems that demonstrate performance or outcomes inconsistent with intended use.","isParaphrase":false},{"shortCode":"BLETCHLEY-2023","name":"Bletchley Declaration on AI Safety","slug":"bletchley-declaration","verdict":"implicit","citation":"Frontier-AI risk frame includes autonomous-action risks; no specific obligation","excerpt":null,"isParaphrase":false},{"shortCode":"SEOUL-2024","name":"Seoul Declaration on Safe, Innovative and Inclusive AI","slug":"seoul-declaration","verdict":"governs","citation":"Frontier AI Safety Commitments §3 — pre-deployment capability evaluations include agentic behaviours under 'realistic deployment conditions'","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF-GENAI","name":"NIST AI RMF Generative AI Profile","slug":"nist-ai-rmf-genai-profile","verdict":"governs","citation":"NIST AI 600-1 names Value Chain + Component Integration as risk category covering agentic / tool-use deployments","excerpt":null,"isParaphrase":false},{"shortCode":"BR-AIBILL-2024","name":"Brazil AI Bill (PL 2338/2023)","slug":"brazil-ai-bill","verdict":"implicit","citation":"Risk-based framework (PL 2338 Arts. 13-15) covers agent systems under high-risk tiers if applicable","excerpt":null,"isParaphrase":false},{"shortCode":"ANTHROPIC-RSP-2024","name":"Anthropic Responsible Scaling Policy (RSP) v2","slug":"anthropic-rsp","verdict":"governs","citation":"RSP v2 — ASL thresholds include 'autonomous AI replication' + agentic capability evaluations","excerpt":null,"isParaphrase":false},{"shortCode":"OPENAI-PREPAREDNESS-2023","name":"OpenAI Preparedness Framework","slug":"openai-preparedness","verdict":"governs","citation":"Preparedness Framework — Model Autonomy is one of four named risk categories","excerpt":null,"isParaphrase":false},{"shortCode":"DEEPMIND-FSF-2024","name":"Google DeepMind Frontier Safety Framework","slug":"deepmind-fsf","verdict":"governs","citation":"FSF Critical Capability Levels — Autonomy is one of four named CCL domains","excerpt":null,"isParaphrase":false},{"shortCode":"META-FRONTIER-2024","name":"Meta Frontier AI Framework","slug":"meta-frontier-ai-framework","verdict":"implicit","citation":"Capability tiers cover agentic behaviour; not named as a distinct category","excerpt":null,"isParaphrase":false},{"shortCode":"UK-US-AISI-MOU-2024","name":"UK-US AI Safety Institute Memorandum of Understanding","slug":"uk-us-aisi-mou","verdict":"implicit","citation":"Joint AISI capability evaluations include agentic-behaviour testing","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-53","name":"California SB-53: Transparency in Frontier Artificial Intelligence Act (TFAIA)","slug":"ca-sb-53","verdict":"implicit","citation":"Bus. & Prof. Code § 22757.11 catastrophic-risk prongs cover a model acting 'without meaningful human oversight' or 'evading the control of its developer or user' (§ 22757.13 incident reporting); reached only via the catastrophic-risk lens, not a dedicated agentic-autonomy regime","excerpt":null,"isParaphrase":false},{"shortCode":"EU-PLD-2024","name":"Revised Product Liability Directive (Directive (EU) 2024/2853)","slug":"eu-product-liability-directive","verdict":"implicit","citation":"Art. 7(2)(c) — defectiveness accounts for a product's ability to continue to learn or acquire new features after market placement; Art. 11(2) — post-placement software-update liability within the manufacturer's control","excerpt":null,"isParaphrase":false},{"shortCode":"EU-PWD-2024","name":"Directive (EU) 2024/2831 on improving working conditions in platform work","slug":"eu-platform-work-directive","verdict":"implicit","citation":"Directive (EU) 2024/2831, Articles 9-11","excerpt":"Automated decision-making systems that autonomously allocate tasks, set pay, monitor and discipline platform workers function as agentic management tools; the Directive subjects them to operative tran","isParaphrase":true},{"shortCode":"NY-RAISE-2025","name":"New York RAISE Act: Responsible AI Safety and Education Act","slug":"ny-raise-act","verdict":"implicit","citation":"N.Y. Gen. Bus. Law § 1420(7) critical harm includes model conduct 'with no meaningful human intervention'; § 1420(13) 'safety incident' includes autonomous model behaviour + control failures — autonomy reached via the catastrophic-risk/incident lens, not a dedicated agentic regime","excerpt":null,"isParaphrase":false}],"governsCount":5,"relatedConcepts":[{"code":"chain-of-thought-monitoring","label":"Chain-of-Thought Monitoring","slug":"chain-of-thought-monitoring"}],"engagingArticles":["agentic-systems-governance","anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","chain-of-thought-monitoring","china-genai-measures","coe-ai-convention","deepmind-fsf","eu-ai-act","eu-platform-work-directive","eu-product-liability-directive","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","ny-raise-act","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"literatureCount":11,"convergence":[{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["agentic-systems-governance","anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","eu-product-liability-directive","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":14},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","eu-ai-act","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":13},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["agentic-systems-governance","anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","eu-product-liability-directive","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","uk-us-aisi-mou"],"citeCount":13},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","eu-ai-act","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":12},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["anthropic-rsp","ca-sb-53","deepmind-fsf","eu-product-liability-directive","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":11},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-53","deepmind-fsf","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":11},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-53","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":8},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["anthropic-rsp","brazil-ai-bill","china-genai-measures","eu-ai-act","g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":7},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["bletchley-declaration","china-genai-measures","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","uk-us-aisi-mou"],"citeCount":6},{"key":"lit-maksym-andriushchenko-alexandra-souly-mateusz-dziemian","title":"AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents","citingArticles":["agentic-systems-governance","bletchley-declaration","deepmind-fsf","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":6},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-53","eu-ai-act","openai-preparedness","seoul-declaration"],"citeCount":6},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["bletchley-declaration","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":6},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["china-genai-measures","g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":4},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["brazil-ai-bill","coe-ai-convention","eu-product-liability-directive","nist-ai-rmf"],"citeCount":4},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":3},{"key":"lit-david-m-allison-stephen-herzog-artificial-intelligence","title":"Artificial Intelligence and Nuclear Weapons Proliferation: The Technological Arms Race for (In)visibility","citingArticles":["bletchley-declaration","meta-frontier-ai-framework","openai-preparedness"],"citeCount":3},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":3},{"key":"lit-tobin-south-samuele-marro-thomas-hardjono-robert-mahar","title":"Authenticated Delegation and Authorized AI Agents","citingArticles":["agentic-systems-governance","eu-product-liability-directive","uk-us-aisi-mou"],"citeCount":3},{"key":"arxiv:2504.18236","title":null,"citingArticles":["coe-ai-convention","nist-ai-rmf"],"citeCount":2},{"key":"lit-alan-chan-carson-ezell-max-kaufmann-kevin-wei-lewis-ha","title":"Visibility into AI Agents","citingArticles":["agentic-systems-governance","eu-product-liability-directive"],"citeCount":2},{"key":"lit-ana-maria-corr-a-sara-garsia-abdullah-elbi-better-toge","title":"Better together? Human oversight as means to achieve fairness in the European AI Act governance","citingArticles":["agentic-systems-governance","uk-us-aisi-mou"],"citeCount":2},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile"],"citeCount":2},{"key":"lit-huijuan-peng-and-pey-woan-lee-reimagining-u-s-tort-law","title":"Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore","citingArticles":["nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":2},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["brazil-ai-bill","coe-ai-convention"],"citeCount":2},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["brazil-ai-bill","china-genai-measures"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["openai-preparedness","uk-us-aisi-mou"],"citeCount":2},{"key":"lit-megha-shrivastava-and-amrita-jash-china-s-semiconducto","title":"China's semiconductor conundrum: understanding US export controls and their efficacy","citingArticles":["anthropic-rsp","meta-frontier-ai-framework"],"citeCount":2},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["china-genai-measures","nist-ai-rmf"],"citeCount":2},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile"],"citeCount":2}]},{"topicCode":"criminal_justice","label":"AI in Criminal Justice","url":"https://policywindow.org/wiki/synthesis/criminal-justice","themeArticle":"https://policywindow.org/wiki/criminal-justice","empiricalConsensus":"contested","contestedQuestion":"Does algorithmic risk-assessment reduce or reproduce racial disparities? Empirical literature (ProPublica COMPAS critique vs. industry replication) is unresolved.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Annex III §6 (high-risk: law enforcement)","excerpt":"Law enforcement: AI systems intended to be used by or on behalf of law enforcement authorities to assess the risk of a natural person offending or re-offending, as polygraphs, or to profile…","isParaphrase":true},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"governs","citation":"§7.1(b) (DOJ AI use review)","excerpt":null,"isParaphrase":false},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"Forensic Information Databases Strategy Board","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"governs","citation":"Art. 14 (procedural safeguards)","excerpt":null,"isParaphrase":false},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"implicit","citation":"Ethical-governance section, paras 62-63 — names law enforcement + the judiciary as sensitive use cases requiring oversight; no dedicated criminal-justice regime","excerpt":null,"isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"governs","citation":"Art. 15 — in judicial use of AI, decisions on legal interpretation/application, evaluation of facts and evidence, and adoption of measures are always reserved to the magistrate; AI limited to organisational/administrative support. Art. 24(2)(h) delegates a future regime for AI in policing.","excerpt":"Nei casi di impiego dei sistemi di intelligenza artificiale nell'attività giudiziaria è sempre riservata al magistrato ogni decisione sull'interpretazione e sull'applicazione della legge, sulla valutazione dei fatti e delle prove e sull'adozione dei provvedimenti.","isParaphrase":false}],"governsCount":4,"relatedConcepts":[],"engagingArticles":["coe-ai-convention","criminal-justice","eu-ai-act","italy-ai-law-2025","uk-ai-white-paper","unesco-ai-ethics-recommendation","us-eo-14110"],"literatureCount":15,"convergence":[{"key":"arxiv:2504.18236","title":null,"citingArticles":["coe-ai-convention","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["eu-ai-act","uk-ai-white-paper","us-eo-14110"],"citeCount":3},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["coe-ai-convention","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["coe-ai-convention","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-shai-farber-machines-of-justice-a-systematic-review-of","title":"Machines of justice: A systematic review of AI applications in policing and criminal justice","citingArticles":["criminal-justice","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2},{"key":"lit-chiara-gallese-predictive-policing-and-predictive-just","title":"Predictive policing and predictive justice: Ethics, data protection, and the AI act","citingArticles":["coe-ai-convention","uk-ai-white-paper"],"citeCount":2},{"key":"lit-emelie-stiernstr-mer-facial-recognition-technology-in","title":"Facial recognition technology in law enforcement: a scoping review of existing empirical studies","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-gary-e-weissman-toni-mankowitz-genevieve-p-kanter-unre","title":"Unregulated large language models produce medical device-like output","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["coe-ai-convention","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["uk-ai-white-paper","us-eo-14110"],"citeCount":2},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["eu-ai-act","uk-ai-white-paper"],"citeCount":2},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2},{"key":"lit-natalie-sheard-algorithm-facilitated-discrimination-a","title":"Algorithm-facilitated discrimination: a socio-legal study of the use by employers of artificial intelligence hiring systems","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2}]},{"topicCode":"education","label":"AI in Education","url":"https://policywindow.org/wiki/synthesis/education","themeArticle":"https://policywindow.org/wiki/education","empiricalConsensus":"settled","contestedQuestion":null,"stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Annex III §3 (high-risk: educational access)","excerpt":"Education and vocational training: AI systems intended to be used to determine access or admission or to assign natural persons to educational and vocational training institutions…","isParaphrase":true},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"implicit","citation":"§8(d) + ED guidance","excerpt":null,"isParaphrase":false},{"shortCode":"UN-RES-2024","name":"UN GA Resolution on Safe, Secure, Trustworthy AI","slug":"un-ai-resolution-2024","verdict":"implicit","citation":"Calls on digital-divide bridging","excerpt":null,"isParaphrase":false},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"governs","citation":"Policy Area 'Education and Research', para 101 — provide adequate AI literacy education to the public","excerpt":"Member States should work with international organizations, educational institutions and private and non-governmental entities to provide adequate AI literacy education to the public","isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"implicit","citation":"No operative schooling regime in force. Art. 24(2)(g) directs (as a delegation criterion) strengthening STEM/artistic competencies in school curricula; Art. 24(2)(i) requires AI-literacy training in universities/AFAM/ITS; Art. 15(4) promotes AI training for magistrates; Art. 22 supports youth.","excerpt":"potenziamento, all'interno dei curricoli scolastici, dello sviluppo di competenze … STEM … attività formative per la comprensione tecnica e l'utilizzo consapevole … dei sistemi di intelligenza artificiale.","isParaphrase":false}],"governsCount":2,"relatedConcepts":[],"engagingArticles":["education","eu-ai-act","italy-ai-law-2025","un-ai-resolution-2024","unesco-ai-ethics-recommendation","us-eo-14110"],"literatureCount":10,"convergence":[{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["eu-ai-act","un-ai-resolution-2024","us-eo-14110"],"citeCount":3},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["un-ai-resolution-2024","us-eo-14110"],"citeCount":2},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2}]},{"topicCode":"employment","label":"AI in Employment","url":"https://policywindow.org/wiki/synthesis/employment","themeArticle":"https://policywindow.org/wiki/employment","empiricalConsensus":"settled","contestedQuestion":null,"stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Annex III §4 (high-risk: employment management)","excerpt":"Employment, workers' management and access to self-employment: AI systems intended to be used for recruitment or selection, and to make decisions affecting terms, promotion, or termination…","isParaphrase":true},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"implicit","citation":"§6 + DOL guidance; sectoral","excerpt":null,"isParaphrase":false},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"ICO + EHRC remit","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"implicit","citation":"Non-discrimination + dignity provisions","excerpt":null,"isParaphrase":false},{"shortCode":"OMB-M-24-10","name":"OMB Memorandum M-24-10 (Advancing Governance, Innovation, and Risk Management for Agency Use of AI)","slug":"omb-m-24-10","verdict":"implicit","citation":"Attachment 1 examples include employment + benefits decisions as rights-impacting; minimum practices apply","excerpt":null,"isParaphrase":false},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"governs","citation":"Policy Area 'Economy and Labour', para 116 — Member States to assess and address AI's impact on labour markets","excerpt":"Member States should assess and address the impact of AI systems on labour markets and its implications for education requirements, in all countries","isParaphrase":false},{"shortCode":"EU-PWD-2024","name":"Directive (EU) 2024/2831 on improving working conditions in platform work","slug":"eu-platform-work-directive","verdict":"governs","citation":"Directive (EU) 2024/2831, Chapter III (esp. Arts. 7-11) and Chapter II (employment-status presumption)","excerpt":"The Directive's core subject is AI in employment: it regulates automated monitoring and decision-making systems used to manage platform workers, requiring human oversight (Art. 10), human review of si","isParaphrase":true},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"governs","citation":"Art. 11 — workplace AI must be safe, reliable, transparent, non-discriminatory and not contrary to human dignity; employer must inform the worker of AI use (per Art. 1-bis D.Lgs. 152/1997). Art. 12 establishes a national Observatory on workplace AI.","excerpt":"L'utilizzo dell'intelligenza artificiale in ambito lavorativo deve essere sicuro, affidabile, trasparente … Il datore di lavoro … è tenuto a informare il lavoratore dell'utilizzo dell'intelligenza artificiale …","isParaphrase":false}],"governsCount":4,"relatedConcepts":[],"engagingArticles":["coe-ai-convention","employment","eu-ai-act","eu-platform-work-directive","italy-ai-law-2025","omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation","us-eo-14110"],"literatureCount":13,"convergence":[{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["coe-ai-convention","omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"lit-natalie-sheard-algorithm-facilitated-discrimination-a","title":"Algorithm-facilitated discrimination: a socio-legal study of the use by employers of artificial intelligence hiring systems","citingArticles":["employment","omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"arxiv:2504.18236","title":null,"citingArticles":["coe-ai-convention","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["eu-ai-act","uk-ai-white-paper","us-eo-14110"],"citeCount":3},{"key":"lit-gary-e-weissman-toni-mankowitz-genevieve-p-kanter-unre","title":"Unregulated large language models produce medical device-like output","citingArticles":["omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["eu-ai-act","omb-m-24-10","us-eo-14110"],"citeCount":3},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["coe-ai-convention","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-sandra-fredman-darcy-du-toit-alessio-bertolini-jonas-v","title":"Fair Work for Platform Workers: Lessons from the EU Directive and Beyond","citingArticles":["employment","eu-platform-work-directive","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2},{"key":"lit-chiara-gallese-predictive-policing-and-predictive-just","title":"Predictive policing and predictive justice: Ethics, data protection, and the AI act","citingArticles":["coe-ai-convention","uk-ai-white-paper"],"citeCount":2},{"key":"lit-emelie-stiernstr-mer-facial-recognition-technology-in","title":"Facial recognition technology in law enforcement: a scoping review of existing empirical studies","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["coe-ai-convention","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-jeremias-adams-prassl-halefom-abraha-aislinn-kelly-lyt","title":"Regulating algorithmic management: A blueprint","citingArticles":["employment","eu-platform-work-directive"],"citeCount":2},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["uk-ai-white-paper","us-eo-14110"],"citeCount":2},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["eu-ai-act","uk-ai-white-paper"],"citeCount":2},{"key":"lit-shai-farber-machines-of-justice-a-systematic-review-of","title":"Machines of justice: A systematic review of AI applications in policing and criminal justice","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-valerio-de-stefano-and-simon-taes-algorithmic-manageme","title":"Algorithmic management and collective bargaining","citingArticles":["employment","eu-platform-work-directive"],"citeCount":2}]},{"topicCode":"healthcare","label":"AI in Healthcare","url":"https://policywindow.org/wiki/synthesis/healthcare","themeArticle":"https://policywindow.org/wiki/healthcare","empiricalConsensus":"settled","contestedQuestion":null,"stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Annex III §5(a) (high-risk: essential services) + MDR overlap","excerpt":"AI systems intended to be used to evaluate the eligibility of natural persons for essential public assistance benefits and services, including healthcare services…","isParaphrase":true},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"implicit","citation":"§8 + HHS strategy","excerpt":null,"isParaphrase":false},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"MHRA software-as-medical-device","excerpt":null,"isParaphrase":false},{"shortCode":"OMB-M-24-10","name":"OMB Memorandum M-24-10 (Advancing Governance, Innovation, and Risk Management for Agency Use of AI)","slug":"omb-m-24-10","verdict":"implicit","citation":"Attachment 1 examples include healthcare access decisions as rights-impacting; minimum practices apply","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-243","name":"California SB 243: Companion Chatbots","slug":"ca-sb-243","verdict":"implicit","citation":"Cal. Bus. & Prof. Code § 22602(b) (added by SB 243) — operator must maintain a protocol preventing production of suicidal-ideation/self-harm content and referring the user to crisis-service providers, published on its website; § 22603 reports referral data to the Office of Suicide Prevention","excerpt":"an operator must maintain a protocol for preventing the production of suicidal ideation, suicide, or self-harm content, including providing a notification that refers the user to crisis service providers when the user expresses such ideation","isParaphrase":true},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"governs","citation":"Policy Area 'Health and Social Well-being', para 121 — employ effective AI for health and the right to life","excerpt":"Member States should endeavour to employ effective AI systems for improving human health and protecting the right to life, including mitigating disease outbreaks","isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"governs","citation":"Art. 7 — AI must not condition access to healthcare on discriminatory criteria (¶2); patient right to be informed of AI use (¶3); the therapeutic decision is always reserved to the physician (¶5). Arts. 8–10 add research, data-processing and electronic-health-record provisions.","excerpt":"L'introduzione di sistemi di intelligenza artificiale nel sistema sanitario non può selezionare e condizionare l'accesso alle prestazioni sanitarie secondo criteri discriminatori. … la decisione … è sempre rimessa agli esercenti la professione medica.","isParaphrase":false}],"governsCount":3,"relatedConcepts":[{"code":"hallucination","label":"Hallucination","slug":"hallucination"}],"engagingArticles":["ca-sb-243","eu-ai-act","hallucination","healthcare","italy-ai-law-2025","omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation","us-eo-14110"],"literatureCount":13,"convergence":[{"key":"lit-gary-e-weissman-toni-mankowitz-genevieve-p-kanter-unre","title":"Unregulated large language models produce medical device-like output","citingArticles":["ca-sb-243","hallucination","healthcare","omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":6},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["ca-sb-243","hallucination","omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"arxiv:2504.18236","title":null,"citingArticles":["hallucination","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-aditya-loganathan-michael-friedman-tayab-waseem-et-al","title":"Current state of Food and Drug Administration-approved artificial intelligence/machine learning medical devices: pathways, transparency, and evidence gaps","citingArticles":["hallucination","healthcare","uk-ai-white-paper"],"citeCount":3},{"key":"lit-boris-babic-i-glenn-cohen-ariel-dora-stern-yiwen-li-me","title":"A general framework for governing marketed AI/ML medical devices","citingArticles":["ca-sb-243","hallucination","healthcare"],"citeCount":3},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["eu-ai-act","uk-ai-white-paper","us-eo-14110"],"citeCount":3},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["eu-ai-act","omb-m-24-10","us-eo-14110"],"citeCount":3},{"key":"lit-natalie-sheard-algorithm-facilitated-discrimination-a","title":"Algorithm-facilitated discrimination: a socio-legal study of the use by employers of artificial intelligence hiring systems","citingArticles":["omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-vijaytha-muralidharan-madelena-y-ng-shada-alsalamah-sa","title":"Global Initiative on AI for Health (GI-AI4H): strategic priorities advancing governance across the United Nations","citingArticles":["ca-sb-243","hallucination","healthcare"],"citeCount":3},{"key":"concept-hallucination","title":"Hallucination","citingArticles":["hallucination","healthcare"],"citeCount":2},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2},{"key":"lit-emelie-stiernstr-mer-facial-recognition-technology-in","title":"Facial recognition technology in law enforcement: a scoping review of existing empirical studies","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["uk-ai-white-paper","us-eo-14110"],"citeCount":2},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["eu-ai-act","uk-ai-white-paper"],"citeCount":2},{"key":"lit-oscar-freyer-isabella-catharina-wiest-jakob-nikolas-ka","title":"A future role for health applications of large language models depends on regulators enforcing safety standards","citingArticles":["ca-sb-243","healthcare"],"citeCount":2},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-sarah-sterz-kevin-baum-sebastian-biewer-holger-hermann","title":"On the Quest for Effectiveness in Human Oversight: Interdisciplinary Perspectives","citingArticles":["ca-sb-243","hallucination"],"citeCount":2},{"key":"lit-shai-farber-machines-of-justice-a-systematic-review-of","title":"Machines of justice: A systematic review of AI applications in policing and criminal justice","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2}]},{"topicCode":"ai_worker_displacement","label":"AI-Driven Worker Displacement","url":"https://policywindow.org/wiki/synthesis/ai-worker-displacement","themeArticle":"https://policywindow.org/wiki/ai-worker-displacement","empiricalConsensus":"emerging","contestedQuestion":"Should displacement governance attach to (a) AI providers (originator liability), (b) AI deployers (use-context liability), or (c) state-level retraining + transition programmes (collectivised response)? Each regime allocates the transition burden differently.","stances":[{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"implicit","citation":"§6 workforce + §6(c) future-of-work studies; not operational obligations","excerpt":null,"isParaphrase":false},{"shortCode":"OECD-AI-PRIN","name":"OECD AI Principles (Recommendation)","slug":"oecd-ai-principles","verdict":"implicit","citation":"Principle 1.1 inclusive growth; OECD AI + Recommendation on AI in workforce (separate instrument)","excerpt":null,"isParaphrase":false},{"shortCode":"UN-RES-2024","name":"UN GA Resolution on Safe, Secure, Trustworthy AI","slug":"un-ai-resolution-2024","verdict":"implicit","citation":"SDG references include decent work + economic growth","excerpt":null,"isParaphrase":false},{"shortCode":"BR-AIBILL-2024","name":"Brazil AI Bill (PL 2338/2023)","slug":"brazil-ai-bill","verdict":"implicit","citation":"PL 2338 has explicit worker-rights provisions + just-transition framing distinctive vs EU AIA","excerpt":null,"isParaphrase":false},{"shortCode":"AU-AI-STRATEGY-2024","name":"African Union Continental AI Strategy","slug":"au-continental-ai-strategy","verdict":"implicit","citation":"Continental strategy includes capacity-building + economic transformation themes that touch displacement","excerpt":null,"isParaphrase":false},{"shortCode":"JP-METI-AI-2024","name":"Japan METI AI Guidelines for Business","slug":"japan-meti-ai-guidelines","verdict":"implicit","citation":"Principle 7 fair competition + workforce themes brush against displacement","excerpt":null,"isParaphrase":false},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"implicit","citation":"Policy Area 'Economy and Labour', para 118 — fair transition (upskilling/reskilling) for at-risk workers; a sub-provision of the labour area","excerpt":null,"isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"implicit","citation":"Art. 12 establishes a national Observatory on the adoption of AI in the workplace charged with study, monitoring and technical support on the occupational, organisational and training effects of AI; Art. 11(1) frames AI as improving working conditions and productivity. Monitoring, not displacement protection.","excerpt":"[Osservatorio sull'adozione di sistemi di intelligenza artificiale nel mondo del lavoro] — [study/monitoring of the occupational, organisational and training effects of AI].","isParaphrase":true},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"implicit","citation":"GDC Objective 5 narrative (A/RES/79/1, Annex I)","excerpt":"efforts to address potential negative impacts of emerging digital technologies on labour and employment.","isParaphrase":false}],"governsCount":0,"relatedConcepts":[],"engagingArticles":["ai-worker-displacement","au-continental-ai-strategy","brazil-ai-bill","italy-ai-law-2025","japan-meti-ai-guidelines","oecd-ai-principles","un-ai-resolution-2024","un-global-digital-compact","unesco-ai-ethics-recommendation","us-eo-14110"],"literatureCount":19,"convergence":[{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["ai-worker-displacement","au-continental-ai-strategy","japan-meti-ai-guidelines","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":6},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["ai-worker-displacement","au-continental-ai-strategy","brazil-ai-bill","japan-meti-ai-guidelines","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":6},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["au-continental-ai-strategy","brazil-ai-bill","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["au-continental-ai-strategy","brazil-ai-bill","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["au-continental-ai-strategy","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["brazil-ai-bill","japan-meti-ai-guidelines","un-ai-resolution-2024","us-eo-14110"],"citeCount":4},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["brazil-ai-bill","japan-meti-ai-guidelines","oecd-ai-principles","us-eo-14110"],"citeCount":4},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["brazil-ai-bill","japan-meti-ai-guidelines","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"arxiv:2504.18236","title":null,"citingArticles":["japan-meti-ai-guidelines","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["brazil-ai-bill","japan-meti-ai-guidelines","oecd-ai-principles"],"citeCount":3},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["au-continental-ai-strategy","un-ai-resolution-2024"],"citeCount":2},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["au-continental-ai-strategy","japan-meti-ai-guidelines"],"citeCount":2},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["japan-meti-ai-guidelines","oecd-ai-principles"],"citeCount":2},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["un-ai-resolution-2024","us-eo-14110"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["au-continental-ai-strategy","japan-meti-ai-guidelines"],"citeCount":2},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["japan-meti-ai-guidelines","un-ai-resolution-2024"],"citeCount":2}]},{"topicCode":"biometric_id","label":"Biometric Identification","url":"https://policywindow.org/wiki/synthesis/biometric-id","themeArticle":"https://policywindow.org/wiki/biometric-id","empiricalConsensus":"settled","contestedQuestion":null,"stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Art. 5(1)(h) prohibition + Art. 26(10) post-hoc rules","excerpt":"the use of 'real-time' remote biometric identification systems in publicly accessible spaces for the purposes of law enforcement, unless and in so far as such use is strictly necessary for…","isParaphrase":false},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"implicit","citation":"§7 civil rights; sectoral agencies retain authority","excerpt":null,"isParaphrase":false},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"ICO + Surveillance Camera Commissioner remit","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"implicit","citation":"Arts. 10-11 (privacy + non-discrimination)","excerpt":null,"isParaphrase":false},{"shortCode":"EU-GDPR-2016","name":"General Data Protection Regulation (GDPR)","slug":"gdpr","verdict":"governs","citation":"Art. 9 special-category processing (biometric data for unique identification); Art. 22 ADM with safeguards","excerpt":null,"isParaphrase":false},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"implicit","citation":"Proportionality & do-no-harm principle (AI should not be used for mass surveillance/social scoring) + Right to privacy principle (para 74, biometric data) — no dedicated biometric-ID provision","excerpt":null,"isParaphrase":false},{"shortCode":"EU-PWD-2024","name":"Directive (EU) 2024/2831 on improving working conditions in platform work","slug":"eu-platform-work-directive","verdict":"governs","citation":"Directive (EU) 2024/2831, Article 7","excerpt":"Article 7 prohibits digital labour platforms from processing biometric data of persons performing platform work to establish identity by one-to-many comparison against a database, while permitting one","isParaphrase":true},{"shortCode":"CN-DEEPSYN-2022","name":"Provisions on the Administration of Deep Synthesis of Internet Information Services","slug":"china-deep-synthesis-provisions","verdict":"governs","citation":"Art. 14","excerpt":"深度合成服务提供者和技术支持者提供人脸、人声等生物识别信息编辑功能的，应当提示深度合成服务使用者依法告知被编辑的个人，并取得其单独同意。","isParaphrase":false}],"governsCount":4,"relatedConcepts":[],"engagingArticles":["biometric-id","china-deep-synthesis-provisions","coe-ai-convention","eu-ai-act","eu-platform-work-directive","gdpr","uk-ai-white-paper","unesco-ai-ethics-recommendation","us-eo-14110"],"literatureCount":13,"convergence":[{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["coe-ai-convention","gdpr","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["biometric-id","coe-ai-convention","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"arxiv:2504.18236","title":null,"citingArticles":["coe-ai-convention","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["eu-ai-act","uk-ai-white-paper","us-eo-14110"],"citeCount":3},{"key":"lit-emelie-stiernstr-mer-facial-recognition-technology-in","title":"Facial recognition technology in law enforcement: a scoping review of existing empirical studies","citingArticles":["biometric-id","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2},{"key":"lit-chiara-gallese-predictive-policing-and-predictive-just","title":"Predictive policing and predictive justice: Ethics, data protection, and the AI act","citingArticles":["coe-ai-convention","uk-ai-white-paper"],"citeCount":2},{"key":"lit-gary-e-weissman-toni-mankowitz-genevieve-p-kanter-unre","title":"Unregulated large language models produce medical device-like output","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["coe-ai-convention","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["uk-ai-white-paper","us-eo-14110"],"citeCount":2},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["eu-ai-act","uk-ai-white-paper"],"citeCount":2},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2},{"key":"lit-natalie-sheard-algorithm-facilitated-discrimination-a","title":"Algorithm-facilitated discrimination: a socio-legal study of the use by employers of artificial intelligence hiring systems","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-novelli-casolari-hacker-spedicato-floridi-generative-a","title":"Generative AI in EU law: Liability, privacy, intellectual property, and cybersecurity","citingArticles":["china-deep-synthesis-provisions","gdpr"],"citeCount":2},{"key":"lit-sandra-fredman-darcy-du-toit-alessio-bertolini-jonas-v","title":"Fair Work for Platform Workers: Lessons from the EU Directive and Beyond","citingArticles":["eu-platform-work-directive","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-shai-farber-machines-of-justice-a-systematic-review-of","title":"Machines of justice: A systematic review of AI applications in policing and criminal justice","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-wachter-mittelstadt-floridi-why-a-right-to-explanation","title":"Why a Right to Explanation of Automated Decision-Making Does Not Exist in the General Data Protection Regulation","citingArticles":["eu-platform-work-directive","gdpr"],"citeCount":2}]},{"topicCode":"catastrophic_risk","label":"Catastrophic & Existential Risk","url":"https://policywindow.org/wiki/synthesis/catastrophic-risk","themeArticle":"https://policywindow.org/wiki/catastrophic-risk","empiricalConsensus":"contested","contestedQuestion":"Are current frontier-model capabilities a meaningful contribution to catastrophic-risk probability? Field is split between catastrophic-risk-as-imminent (FLI, CAIS) and catastrophic-risk-as-speculative (Pope et al., Andersson) positions.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"implicit","citation":"Art. 51 + Recital 32 — systemic risk overlaps with but does not fully cover catastrophic-risk framing","excerpt":"'systemic risk' means a risk specific to the high-impact capabilities of general-purpose AI models … with significant impact on the Union market or on public health, safety, security, or fundamental rights…","isParaphrase":true},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"governs","citation":"§4.2(a)(ii) — CBRN + autonomous replication explicitly named","excerpt":null,"isParaphrase":false},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"AISI remit covers frontier-model evaluation; not in white paper text","excerpt":null,"isParaphrase":false},{"shortCode":"G7-HIROSHIMA","name":"G7 Hiroshima AI Process Code of Conduct","slug":"g7-hiroshima-code","verdict":"governs","citation":"Code §1 + §3 — explicit risk-identification including CBRN","excerpt":null,"isParaphrase":false},{"shortCode":"UN-RES-2024","name":"UN GA Resolution on Safe, Secure, Trustworthy AI","slug":"un-ai-resolution-2024","verdict":"implicit","citation":"Notes 'shared concerns' but no operative catastrophic-risk text","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF","name":"NIST AI Risk Management Framework","slug":"nist-ai-rmf","verdict":"implicit","citation":"Map 1.1 risk classification covers catastrophic via 'societal' impact tier; GenAI Profile (2024) adds explicit content","excerpt":"Processes, procedures, and practices are in place to determine the needed level of risk management activities based on the organization's risk tolerance.","isParaphrase":false},{"shortCode":"BLETCHLEY-2023","name":"Bletchley Declaration on AI Safety","slug":"bletchley-declaration","verdict":"governs","citation":"Declaration §3-5 (substantial risks from frontier AI, including catastrophic harm)","excerpt":null,"isParaphrase":false},{"shortCode":"SEOUL-2024","name":"Seoul Declaration on Safe, Innovative and Inclusive AI","slug":"seoul-declaration","verdict":"governs","citation":"Frontier AI Safety Commitments §1: identify thresholds for severe risks pre-deployment","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF-GENAI","name":"NIST AI RMF Generative AI Profile","slug":"nist-ai-rmf-genai-profile","verdict":"governs","citation":"NIST AI 600-1 §3.1 CBRN Information Uplift; §3.3 Dangerous, Violent, or Hateful Content","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-1047","name":"California SB-1047: Safe and Secure Innovation for Frontier AI Models Act","slug":"ca-sb-1047","verdict":"governs","citation":"Cal. SB-1047 §22602 — defines 'critical harm' including mass casualties, $500M+ damage","excerpt":null,"isParaphrase":false},{"shortCode":"BR-AIBILL-2024","name":"Brazil AI Bill (PL 2338/2023)","slug":"brazil-ai-bill","verdict":"governs","citation":"PL 2338/2023 Art. 14 (excessive-risk AI applications — explicit prohibition + risk-tier framework)","excerpt":null,"isParaphrase":false},{"shortCode":"ANTHROPIC-RSP-2024","name":"Anthropic Responsible Scaling Policy (RSP) v2","slug":"anthropic-rsp","verdict":"governs","citation":"RSP v2 §3 — ASL-3 / ASL-4 capability thresholds explicitly target CBRN uplift + autonomous-replication","excerpt":null,"isParaphrase":false},{"shortCode":"OPENAI-PREPAREDNESS-2023","name":"OpenAI Preparedness Framework","slug":"openai-preparedness","verdict":"governs","citation":"Preparedness Framework risk-tier matrix — Critical tier explicitly targets CBRN, cyber, persuasion, autonomy","excerpt":null,"isParaphrase":false},{"shortCode":"DEEPMIND-FSF-2024","name":"Google DeepMind Frontier Safety Framework","slug":"deepmind-fsf","verdict":"governs","citation":"FSF Critical Capability Levels (CCL) — explicit thresholds for autonomy, biosecurity, cyber, persuasion","excerpt":null,"isParaphrase":false},{"shortCode":"META-FRONTIER-2024","name":"Meta Frontier AI Framework","slug":"meta-frontier-ai-framework","verdict":"governs","citation":"Framework critical-risk tier — commit to halt training pre-mitigation if reached","excerpt":null,"isParaphrase":false},{"shortCode":"UK-US-AISI-MOU-2024","name":"UK-US AI Safety Institute Memorandum of Understanding","slug":"uk-us-aisi-mou","verdict":"implicit","citation":"Joint evaluation scope encompasses CBRN + autonomy uplift questions; MoU text does not enumerate explicit thresholds","excerpt":null,"isParaphrase":false},{"shortCode":"WH-VOLUNTARY-2023","name":"White House Voluntary AI Commitments","slug":"wh-voluntary-2023","verdict":"implicit","citation":"Commitments §1 references CBRN + bio risks via 'most significant societal risks'; not threshold-explicit","excerpt":null,"isParaphrase":false},{"shortCode":"EU-GPAI-COP-2025","name":"EU General-Purpose AI Code of Practice","slug":"gpai-code-of-practice","verdict":"governs","citation":"Chapter 3 systemic-risk-tier capability evaluations + serious-incident reporting + model-weight access controls (Art. 55 substrate)","excerpt":null,"isParaphrase":false},{"shortCode":"DOD-RAI-2022","name":"DoD Responsible AI Strategy and Implementation Pathway","slug":"dod-rai-strategy","verdict":"implicit","citation":"Ethical Principle 'Reliable' + Tenet 4 (Requirements Validation) — JCIDS gating addresses mission-risk; DoDD 3000.09 separately governs autonomy-in-weapons LAWS-specific catastrophic-risk decisions","excerpt":"The Department's AI capabilities will have explicit, well-defined uses, and the safety, security, and effectiveness of such capabilities will be subject to testing and assurance within those defined uses…","isParaphrase":false},{"shortCode":"CA-SB-53","name":"California SB-53: Transparency in Frontier Artificial Intelligence Act (TFAIA)","slug":"ca-sb-53","verdict":"governs","citation":"Bus. & Prof. Code § 22757.11 (definition) operationalized by §§ 22757.12 (framework) + 22757.13 (critical-safety-incident reporting to CalOES)","excerpt":"'Catastrophic risk' means a foreseeable and material risk that a frontier developer's … frontier model will materially contribute to the death of, or serious injury to, more than 50 people or more than one billion dollars ($1,000,000,000) in damage to, or loss of, property…","isParaphrase":true},{"shortCode":"NY-RAISE-2025","name":"New York RAISE Act: Responsible AI Safety and Education Act","slug":"ny-raise-act","verdict":"governs","citation":"N.Y. Gen. Bus. Law § 1421(1) requires a large developer to implement and conspicuously publish a written safety and security protocol governing the risk of 'critical harm' from its frontier models, and § 1421(4) requires disclosure of safety incidents within 72 hours; § 1420(7) defines critical harm (100+ deaths/serious injuries or $1B damage via CBRN weapons or autonomous model conduct). NOTE: the floor-text § 1421(2) deployment PROHIBITION was struck by the chapter amendment enacted Mar. 27, 2026 (S8828/A9449), which reoriented the Act to a transparency-and-reporting regime; this cell tracks the RETAINED safety-protocol + incident-reporting duties, not a deployment ban.","excerpt":"A large developer shall implement a written safety and security protocol [addressing the risk of critical harm] and conspicuously publish it with appropriate redactions, transmitting a copy to the attorney general.","isParaphrase":true},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"implicit","citation":"GDC Objective 5, paras 55(a) and 56(a) (A/RES/79/1, Annex I)","excerpt":"Assess the future directions and implications of artificial intelligence systems ... evidence-based impact, risk and opportunity assessments.","isParaphrase":true}],"governsCount":14,"relatedConcepts":[{"code":"agentic-system","label":"Agentic AI System","slug":"agentic-system"},{"code":"tool-use-safety","label":"Tool-Use Safety","slug":"tool-use-safety"},{"code":"jailbreak-resistance","label":"Jailbreak Resistance","slug":"jailbreak-resistance"},{"code":"chain-of-thought-monitoring","label":"Chain-of-Thought Monitoring","slug":"chain-of-thought-monitoring"}],"engagingArticles":["agentic-system","anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-1047","ca-sb-53","catastrophic-risk","chain-of-thought-monitoring","deepmind-fsf","dod-rai-strategy","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","jailbreak-resistance","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","ny-raise-act","openai-preparedness","seoul-declaration","tool-use-safety","uk-ai-white-paper","uk-us-aisi-mou","un-ai-resolution-2024","un-global-digital-compact","us-eo-14110","wh-voluntary-2023"],"literatureCount":18,"convergence":[{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-1047","ca-sb-53","deepmind-fsf","dod-rai-strategy","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","tool-use-safety","uk-ai-white-paper","uk-us-aisi-mou","us-eo-14110","wh-voluntary-2023"],"citeCount":20},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-1047","ca-sb-53","deepmind-fsf","dod-rai-strategy","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","meta-frontier-ai-framework","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","tool-use-safety","uk-ai-white-paper","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":18},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","deepmind-fsf","g7-hiroshima-code","gpai-code-of-practice","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-ai-white-paper","uk-us-aisi-mou","un-ai-resolution-2024","us-eo-14110","wh-voluntary-2023"],"citeCount":17},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","catastrophic-risk","deepmind-fsf","dod-rai-strategy","gpai-code-of-practice","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","tool-use-safety","uk-us-aisi-mou","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":15},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":12},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","uk-us-aisi-mou"],"citeCount":11},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["anthropic-rsp","ca-sb-53","deepmind-fsf","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":10},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["anthropic-rsp","brazil-ai-bill","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","nist-ai-rmf","nist-ai-rmf-genai-profile","un-ai-resolution-2024","us-eo-14110","wh-voluntary-2023"],"citeCount":10},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","dod-rai-strategy","eu-ai-act","openai-preparedness","seoul-declaration","us-eo-14110","wh-voluntary-2023"],"citeCount":10},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["bletchley-declaration","ca-sb-1047","catastrophic-risk","deepmind-fsf","dod-rai-strategy","gpai-code-of-practice","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":10},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["bletchley-declaration","ca-sb-1047","dod-rai-strategy","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","uk-ai-white-paper","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":9},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":5},{"key":"lit-bengio-hinton-yao-song-et-al-managing-extreme-ai-risks","title":"Managing extreme AI risks amid rapid progress","citingArticles":["ca-sb-1047","catastrophic-risk","dod-rai-strategy","tool-use-safety","uk-us-aisi-mou"],"citeCount":5},{"key":"lit-maksym-andriushchenko-alexandra-souly-mateusz-dziemian","title":"AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents","citingArticles":["bletchley-declaration","deepmind-fsf","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":5},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["brazil-ai-bill","ca-sb-1047","dod-rai-strategy","nist-ai-rmf","uk-ai-white-paper"],"citeCount":5},{"key":"arxiv:2504.18236","title":null,"citingArticles":["ca-sb-1047","dod-rai-strategy","nist-ai-rmf","uk-ai-white-paper"],"citeCount":4},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["gpai-code-of-practice","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":4},{"key":"lit-david-m-allison-stephen-herzog-artificial-intelligence","title":"Artificial Intelligence and Nuclear Weapons Proliferation: The Technological Arms Race for (In)visibility","citingArticles":["bletchley-declaration","dod-rai-strategy","meta-frontier-ai-framework","openai-preparedness"],"citeCount":4},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile","wh-voluntary-2023"],"citeCount":4},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":4},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["g7-hiroshima-code","uk-ai-white-paper","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-huijuan-peng-and-pey-woan-lee-reimagining-u-s-tort-law","title":"Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore","citingArticles":["nist-ai-rmf","nist-ai-rmf-genai-profile","wh-voluntary-2023"],"citeCount":3},{"key":"lit-jonas-schuett-markus-anderljung-alexis-carlier-leonie","title":"From Principles to Rules: A Regulatory Approach for Frontier AI","citingArticles":["catastrophic-risk","jailbreak-resistance","tool-use-safety"],"citeCount":3},{"key":"lit-megha-shrivastava-and-amrita-jash-china-s-semiconducto","title":"China's semiconductor conundrum: understanding US export controls and their efficacy","citingArticles":["anthropic-rsp","meta-frontier-ai-framework","us-eo-14110"],"citeCount":3},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["nist-ai-rmf","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":3},{"key":"lit-anderljung-barnhart-korinek-et-al-frontier-ai-regulati","title":"Frontier AI Regulation: Managing Emerging Risks to Public Safety","citingArticles":["catastrophic-risk","us-eo-14110"],"citeCount":2},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["brazil-ai-bill","un-ai-resolution-2024"],"citeCount":2},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["brazil-ai-bill","un-ai-resolution-2024"],"citeCount":2},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["brazil-ai-bill","un-ai-resolution-2024"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["openai-preparedness","uk-us-aisi-mou"],"citeCount":2},{"key":"lit-mary-phuong-matthew-aitchison-elliot-catt-et-al-google","title":"Evaluating Frontier Models for Dangerous Capabilities","citingArticles":["jailbreak-resistance","tool-use-safety"],"citeCount":2},{"key":"lit-shevlane-farquhar-garfinkel-et-al-model-evaluation-for","title":"Model evaluation for extreme risks","citingArticles":["catastrophic-risk","tool-use-safety"],"citeCount":2}]},{"topicCode":"compute_export_controls","label":"Compute + Model-Weight Export Controls","url":"https://policywindow.org/wiki/synthesis/compute-export-controls","themeArticle":"https://policywindow.org/wiki/compute-export-controls","empiricalConsensus":"contested","contestedQuestion":"Should compute + weight export controls govern by (a) recipient jurisdiction, (b) capability tier of the controlled artifact, or (c) end-use intent? Each rule generation has shifted between these frames.","stances":[{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"implicit","citation":"§4.2(b) directs export-control coordination via BIS; not the primary venue but the policy hook","excerpt":null,"isParaphrase":false},{"shortCode":"ANTHROPIC-RSP-2024","name":"Anthropic Responsible Scaling Policy (RSP) v2","slug":"anthropic-rsp","verdict":"implicit","citation":"ASL-3+ tiers include model-weight access controls (recipient-restriction analog)","excerpt":null,"isParaphrase":false},{"shortCode":"DEEPMIND-FSF-2024","name":"Google DeepMind Frontier Safety Framework","slug":"deepmind-fsf","verdict":"implicit","citation":"FSF mitigations include model-weight access controls + restricted-deployment options","excerpt":null,"isParaphrase":false},{"shortCode":"META-FRONTIER-2024","name":"Meta Frontier AI Framework","slug":"meta-frontier-ai-framework","verdict":"implicit","citation":"Framework's release decisions implicitly determine cross-border weight flow","excerpt":null,"isParaphrase":false}],"governsCount":0,"relatedConcepts":[{"code":"hardware-enabled-governance","label":"Hardware-Enabled Governance Mechanisms","slug":"hardware-enabled-governance"}],"engagingArticles":["anthropic-rsp","compute-export-controls","deepmind-fsf","hardware-enabled-governance","meta-frontier-ai-framework","us-eo-14110"],"literatureCount":9,"convergence":[{"key":"lit-megha-shrivastava-and-amrita-jash-china-s-semiconducto","title":"China's semiconductor conundrum: understanding US export controls and their efficacy","citingArticles":["anthropic-rsp","compute-export-controls","hardware-enabled-governance","meta-frontier-ai-framework","us-eo-14110"],"citeCount":5},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework","us-eo-14110"],"citeCount":4},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework","us-eo-14110"],"citeCount":4},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework"],"citeCount":3},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework"],"citeCount":3},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework"],"citeCount":3},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework"],"citeCount":3},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework"],"citeCount":3},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["anthropic-rsp","us-eo-14110"],"citeCount":2},{"key":"lit-chan-yuan-wong-henry-wai-chung-yeung-shaopeng-huang-ja","title":"Geopolitics and the changing landscape of global value chains and competition in the global semiconductor industry: Rivalry and catch-up in chip manufacturing in East Asia","citingArticles":["compute-export-controls","hardware-enabled-governance"],"citeCount":2},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["hardware-enabled-governance","us-eo-14110"],"citeCount":2},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["anthropic-rsp","us-eo-14110"],"citeCount":2},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["deepmind-fsf","meta-frontier-ai-framework"],"citeCount":2},{"key":"lit-sastry-heim-belfield-anderljung-brundage-et-al-computi","title":"Computing Power and the Governance of Artificial Intelligence","citingArticles":["compute-export-controls","hardware-enabled-governance"],"citeCount":2}]},{"topicCode":"compute_reporting","label":"Compute-Threshold Reporting","url":"https://policywindow.org/wiki/synthesis/compute-reporting","themeArticle":"https://policywindow.org/wiki/compute-reporting","empiricalConsensus":"contested","contestedQuestion":"Are compute thresholds (10²⁵ FLOPs EU, 10²⁶ FLOPs US) a defensible proxy for governance-relevant capability, given algorithmic-efficiency improvements? Field is split.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Art. 51(2) + Annex XIII (10²⁵ FLOP presumption)","excerpt":"a general-purpose AI model shall be presumed to have high impact capabilities … when the cumulative amount of computation used for its training measured in floating point operations is greater than 10^25.","isParaphrase":true},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"governs","citation":"§4.2(a)(i) — 10²⁶ FLOP threshold","excerpt":null,"isParaphrase":false},{"shortCode":"BLETCHLEY-2023","name":"Bletchley Declaration on AI Safety","slug":"bletchley-declaration","verdict":"implicit","citation":"Declaration §6 calls for capability evaluation but does not specify compute thresholds","excerpt":null,"isParaphrase":false},{"shortCode":"SEOUL-2024","name":"Seoul Declaration on Safe, Innovative and Inclusive AI","slug":"seoul-declaration","verdict":"implicit","citation":"Safety Commitments invoke capability thresholds; compute is one proxy","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-1047","name":"California SB-1047: Safe and Secure Innovation for Frontier AI Models Act","slug":"ca-sb-1047","verdict":"governs","citation":"Cal. SB-1047 §22603(b) — annual reporting of training compute + safety determination","excerpt":null,"isParaphrase":false},{"shortCode":"ANTHROPIC-RSP-2024","name":"Anthropic Responsible Scaling Policy (RSP) v2","slug":"anthropic-rsp","verdict":"implicit","citation":"RSP v2 capability evaluations triggered by capability rather than pure compute; compute is one signal","excerpt":null,"isParaphrase":false},{"shortCode":"OPENAI-PREPAREDNESS-2023","name":"OpenAI Preparedness Framework","slug":"openai-preparedness","verdict":"implicit","citation":"Capability-tier evaluations are the primary trigger; compute is a coincident signal","excerpt":null,"isParaphrase":false},{"shortCode":"WH-VOLUNTARY-2023","name":"White House Voluntary AI Commitments","slug":"wh-voluntary-2023","verdict":"implicit","citation":"Self-reporting through commitments framework; binding compute thresholds came via EO 14110 §4.2(a)","excerpt":null,"isParaphrase":false},{"shortCode":"OMB-M-24-10","name":"OMB Memorandum M-24-10 (Advancing Governance, Innovation, and Risk Management for Agency Use of AI)","slug":"omb-m-24-10","verdict":"governs","citation":"§3(a)(iv)–(v) annual public AI use-case inventory + quarterly AI procurement reporting to OMB","excerpt":"Agencies must report to OMB and, as appropriate, publicly release aggregate metrics about their AI use cases that are determined to be safety-impacting or rights-impacting.","isParaphrase":true},{"shortCode":"GSA-AI-GUIDE-2024","name":"GSA Generative AI and Specialized Computing Infrastructure Acquisition Resource Guide","slug":"gsa-ai-acquisition-guide","verdict":"governs","citation":"Guide routes AI acquisitions through existing governmentwide vehicles (MAS IT / Best-in-Class GWACs) rather than a dedicated generative-AI vehicle or new AI-specific SINs","excerpt":"Faithful summary: the guide routes AI acquisitions through existing governmentwide vehicles (MAS IT and Best-in-Class GWACs), noting there is no generative-AI-only vehicle; it does not enumerate new AI-specific Special Item Numbers.","isParaphrase":true},{"shortCode":"DOD-RAI-2022","name":"DoD Responsible AI Strategy and Implementation Pathway","slug":"dod-rai-strategy","verdict":"implicit","citation":"Tenet 1 (RAI Governance) + Tenet 3 (Acquisition Lifecycle) — clarifies CDAO + OUSD(A&S) roles in AI procurement oversight; tracking + reporting emerge through standard DoD acquisition reporting channels","excerpt":null,"isParaphrase":false},{"shortCode":"FEDRAMP-AI-2024","name":"FedRAMP AI Cloud Procurement Guidance","slug":"fedramp-ai-guidance","verdict":"implicit","citation":"FedRAMP authorisation enables ATO; agency-AI-use disclosure flows through OMB M-24-10 inventory + quarterly procurement reporting rather than through FedRAMP itself","excerpt":null,"isParaphrase":false},{"shortCode":"DFARS-252-204","name":"DFARS Subpart 252.204 (Safeguarding Covered Defense Information and Cyber Incident Reporting)","slug":"dfars-252-204","verdict":"implicit","citation":"Cyber-incident reporting under 252.204-7012(c) — 72-hour DoD notification covers AI-system compromise events including model-weight theft + prompt-injection-based credential exposure; broader AI-use disclosure flows through M-24-10 not DFARS","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-53","name":"California SB-53: Transparency in Frontier Artificial Intelligence Act (TFAIA)","slug":"ca-sb-53","verdict":"implicit","citation":"Bus. & Prof. Code § 22757.11 uses a 10^26 FLOP compute threshold to SCOPE the regulated class + § 22757.12 ties disclosure to compute-defined frontier models; no standalone compute-figure reporting mandate to a regulator","excerpt":null,"isParaphrase":false},{"shortCode":"NY-RAISE-2025","name":"New York RAISE Act: Responsible AI Safety and Education Act","slug":"ny-raise-act","verdict":"implicit","citation":"N.Y. Gen. Bus. Law § 1420(6),(9) — the frontier-model / large-developer compute figures SCOPE the regulated class; no standalone compute-figure reporting duty to a regulator. (The Mar. 27, 2026 chapter amendment revised the large-developer threshold to align more closely with California's criteria; the verdict — coverage-scoping, not a reporting duty — is unchanged by the specific figure.)","excerpt":null,"isParaphrase":false},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"implicit","citation":"Act No. 53 of 2025, Art. 12","excerpt":"... facilities and equipment relating to large-scale information processing ... the State shall take measures to develop, improve, and promote the shared use of such facilities ...","isParaphrase":true}],"governsCount":5,"relatedConcepts":[{"code":"frontier-tier","label":"Frontier-Tier AI","slug":"frontier-tier"},{"code":"systemic-risk","label":"Systemic Risk (AI)","slug":"systemic-risk"},{"code":"designated-systemic","label":"Designated Systemic-Risk Model","slug":"designated-systemic"},{"code":"compute-threshold","label":"Compute Threshold (AI Governance)","slug":"compute-threshold"},{"code":"red-team-evaluation","label":"Red-Team Evaluation","slug":"red-team-evaluation"},{"code":"alignment","label":"AI Alignment","slug":"alignment"},{"code":"deceptive-alignment","label":"Deceptive Alignment","slug":"deceptive-alignment"},{"code":"mesa-optimization","label":"Mesa-Optimization","slug":"mesa-optimization"},{"code":"capability-elicitation","label":"Capability Elicitation","slug":"capability-elicitation"},{"code":"policy-instrument","label":"Policy Instrument","slug":"policy-instrument"},{"code":"ai-supply-chain","label":"AI Supply Chain","slug":"ai-supply-chain"},{"code":"multi-turn-evaluation","label":"Multi-Turn Evaluation","slug":"multi-turn-evaluation"},{"code":"model-distillation-risk","label":"Model Distillation Risk","slug":"model-distillation-risk"},{"code":"inference-time-compute","label":"Inference-Time Compute","slug":"inference-time-compute"},{"code":"sandbagging","label":"Sandbagging","slug":"sandbagging"},{"code":"in-context-learning","label":"In-Context Learning","slug":"in-context-learning"},{"code":"hardware-enabled-governance","label":"Hardware-Enabled Governance Mechanisms","slug":"hardware-enabled-governance"}],"engagingArticles":["ai-supply-chain","alignment","anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","capability-elicitation","compute-reporting","compute-threshold","deceptive-alignment","designated-systemic","dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","frontier-tier","gsa-ai-acquisition-guide","hardware-enabled-governance","in-context-learning","inference-time-compute","japan-ai-promotion-act","mesa-optimization","model-distillation-risk","multi-turn-evaluation","ny-raise-act","omb-m-24-10","openai-preparedness","policy-instrument","red-team-evaluation","sandbagging","seoul-declaration","systemic-risk","us-eo-14110","wh-voluntary-2023"],"literatureCount":25,"convergence":[{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["ai-supply-chain","anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","capability-elicitation","compute-reporting","compute-threshold","designated-systemic","dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","frontier-tier","gsa-ai-acquisition-guide","in-context-learning","inference-time-compute","model-distillation-risk","omb-m-24-10","openai-preparedness","policy-instrument","red-team-evaluation","seoul-declaration","systemic-risk","us-eo-14110","wh-voluntary-2023"],"citeCount":26},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["ai-supply-chain","anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","capability-elicitation","designated-systemic","dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","frontier-tier","gsa-ai-acquisition-guide","in-context-learning","model-distillation-risk","multi-turn-evaluation","openai-preparedness","policy-instrument","red-team-evaluation","seoul-declaration","systemic-risk","us-eo-14110","wh-voluntary-2023"],"citeCount":23},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","capability-elicitation","designated-systemic","dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide","multi-turn-evaluation","openai-preparedness","red-team-evaluation","seoul-declaration","systemic-risk","wh-voluntary-2023"],"citeCount":17},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["ai-supply-chain","capability-elicitation","compute-reporting","compute-threshold","designated-systemic","dfars-252-204","fedramp-ai-guidance","frontier-tier","gsa-ai-acquisition-guide","hardware-enabled-governance","in-context-learning","inference-time-compute","model-distillation-risk","policy-instrument","systemic-risk","us-eo-14110"],"citeCount":16},{"key":"lit-sastry-heim-belfield-anderljung-brundage-et-al-computi","title":"Computing Power and the Governance of Artificial Intelligence","citingArticles":["ai-supply-chain","ca-sb-1047","capability-elicitation","compute-reporting","compute-threshold","designated-systemic","dfars-252-204","frontier-tier","gsa-ai-acquisition-guide","hardware-enabled-governance","inference-time-compute","model-distillation-risk","omb-m-24-10","policy-instrument","red-team-evaluation","systemic-risk"],"citeCount":16},{"key":"lit-mary-phuong-matthew-aitchison-elliot-catt-et-al-google","title":"Evaluating Frontier Models for Dangerous Capabilities","citingArticles":["capability-elicitation","designated-systemic","frontier-tier","in-context-learning","inference-time-compute","mesa-optimization","model-distillation-risk","multi-turn-evaluation","systemic-risk"],"citeCount":9},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","dod-rai-strategy","openai-preparedness","seoul-declaration","wh-voluntary-2023"],"citeCount":8},{"key":"lit-eloundou-manning-mishkin-rock-gpts-are-gpts-labor-mark","title":"GPTs are GPTs: Labor market impact potential of LLMs","citingArticles":["capability-elicitation","fedramp-ai-guidance","gsa-ai-acquisition-guide","in-context-learning","model-distillation-risk","omb-m-24-10","policy-instrument","red-team-evaluation"],"citeCount":8},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["ai-supply-chain","bletchley-declaration","ca-sb-1047","dfars-252-204","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","wh-voluntary-2023"],"citeCount":8},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","openai-preparedness","seoul-declaration","us-eo-14110","wh-voluntary-2023"],"citeCount":8},{"key":"lit-akash-r-wasil-tom-reed-jack-william-miller-peter-barne","title":"Verification methods for international AI agreements","citingArticles":["capability-elicitation","compute-reporting","compute-threshold","designated-systemic","hardware-enabled-governance","inference-time-compute","model-distillation-risk"],"citeCount":7},{"key":"lit-lehdonvirta-w-hawkins-compute-north-vs-compute-south-t","title":"Compute North vs. Compute South: The Uneven Possibilities of Compute-based AI Governance Around the Globe","citingArticles":["ai-supply-chain","compute-reporting","compute-threshold","designated-systemic","frontier-tier","policy-instrument"],"citeCount":6},{"key":"lit-anderljung-barnhart-korinek-et-al-frontier-ai-regulati","title":"Frontier AI Regulation: Managing Emerging Risks to Public Safety","citingArticles":["capability-elicitation","frontier-tier","multi-turn-evaluation","policy-instrument","us-eo-14110"],"citeCount":5},{"key":"lit-anka-reuel-ben-bucknall-stephen-casper-tim-fist-lennar","title":"Open Problems in Technical AI Governance","citingArticles":["capability-elicitation","compute-threshold","hardware-enabled-governance","in-context-learning","multi-turn-evaluation"],"citeCount":5},{"key":"lit-lennart-heim-tim-fist-janet-egan-sihao-huang-stephen-z","title":"Governing Through the Cloud: The Intermediary Role of Compute Providers in AI Regulation","citingArticles":["compute-reporting","compute-threshold","designated-systemic","inference-time-compute","model-distillation-risk"],"citeCount":5},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["ca-sb-1047","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","omb-m-24-10"],"citeCount":5},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-53","openai-preparedness","seoul-declaration"],"citeCount":5},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["bletchley-declaration","ca-sb-1047","dod-rai-strategy","openai-preparedness","seoul-declaration"],"citeCount":5},{"key":"lit-stephen-weymouth-digital-disintegration-techno-blocs-a","title":"Digital Disintegration: Techno-Blocs and Strategic Sovereignty in the AI Era","citingArticles":["ai-supply-chain","compute-threshold","hardware-enabled-governance","model-distillation-risk","policy-instrument"],"citeCount":5},{"key":"arxiv:2504.18236","title":null,"citingArticles":["ca-sb-1047","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":4},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["anthropic-rsp","ca-sb-53","openai-preparedness","seoul-declaration"],"citeCount":4},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["anthropic-rsp","eu-ai-act","us-eo-14110","wh-voluntary-2023"],"citeCount":4},{"key":"lit-francesca-palmiotto-the-ai-act-roller-coaster-the-evol","title":"The AI Act Roller Coaster: The Evolution of Fundamental Rights Protection in the Legislative Process and the Future of the Regulation","citingArticles":["dfars-252-204","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":4},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-53","openai-preparedness"],"citeCount":4},{"key":"concept-mesa-optimization","title":"Mesa-Optimization","citingArticles":["alignment","deceptive-alignment","mesa-optimization"],"citeCount":3},{"key":"lit-david-m-allison-stephen-herzog-artificial-intelligence","title":"Artificial Intelligence and Nuclear Weapons Proliferation: The Technological Arms Race for (In)visibility","citingArticles":["bletchley-declaration","dod-rai-strategy","openai-preparedness"],"citeCount":3},{"key":"lit-ezgi-yazici-toward-a-global-standard-for-ethical-ai-re","title":"Toward a global standard for ethical AI regulation: addressing gaps in AI-driven biometric and high-resolution satellite imaging in the EU AI Act","citingArticles":["dfars-252-204","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":3},{"key":"lit-justin-kollar-andrew-stokols-geopolitical-ecologies-of","title":"Geopolitical ecologies of cloud capitalism: Territorial restructuring and the making of national computing power in the U.S. and China","citingArticles":["ai-supply-chain","ca-sb-53","compute-threshold"],"citeCount":3},{"key":"lit-maksym-andriushchenko-alexandra-souly-mateusz-dziemian","title":"AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents","citingArticles":["bletchley-declaration","openai-preparedness","seoul-declaration"],"citeCount":3},{"key":"lit-megha-shrivastava-and-amrita-jash-china-s-semiconducto","title":"China's semiconductor conundrum: understanding US export controls and their efficacy","citingArticles":["anthropic-rsp","hardware-enabled-governance","us-eo-14110"],"citeCount":3},{"key":"arxiv:2311.08379","title":null,"citingArticles":["deceptive-alignment","sandbagging"],"citeCount":2},{"key":"arxiv:2407.05694","title":null,"citingArticles":["compute-threshold","systemic-risk"],"citeCount":2},{"key":"arxiv:2412.14093","title":null,"citingArticles":["deceptive-alignment","sandbagging"],"citeCount":2},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["ai-supply-chain","gsa-ai-acquisition-guide"],"citeCount":2},{"key":"lit-bengio-hinton-yao-song-et-al-managing-extreme-ai-risks","title":"Managing extreme AI risks amid rapid progress","citingArticles":["ca-sb-1047","dod-rai-strategy"],"citeCount":2},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["openai-preparedness","policy-instrument"],"citeCount":2},{"key":"lit-henrik-palmer-olsen-thomas-troels-hildebrandt-corneliu","title":"The Right to Transparency in Public Governance: Freedom of Information and the Use of Artificial Intelligence by Public Agencies","citingArticles":["in-context-learning","multi-turn-evaluation"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["openai-preparedness","policy-instrument"],"citeCount":2},{"key":"lit-martin-kretschmer-bartolomeo-meletti-lionel-bently-gab","title":"Copyright and AI in the UK: Opting-In or Opting-Out?","citingArticles":["ai-supply-chain","gsa-ai-acquisition-guide"],"citeCount":2},{"key":"lit-novelli-casolari-hacker-spedicato-floridi-generative-a","title":"Generative AI in EU law: Liability, privacy, intellectual property, and cybersecurity","citingArticles":["ai-supply-chain","dfars-252-204"],"citeCount":2},{"key":"lit-shavit-what-does-it-take-to-catch-a-chinchilla-verifyi","title":"What does it take to catch a Chinchilla? Verifying Rules on Large-Scale Neural Network Training via Compute Monitoring","citingArticles":["compute-reporting","hardware-enabled-governance"],"citeCount":2}]},{"topicCode":"deepfakes","label":"Deepfakes / Synthetic Content","url":"https://policywindow.org/wiki/synthesis/deepfakes","themeArticle":"https://policywindow.org/wiki/deepfakes","empiricalConsensus":"contested","contestedQuestion":"Is robust watermarking durable under adversarial removal at deployment scale? Field is split on technical feasibility despite policy convergence on the requirement.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Art. 50(4) (disclosure obligation for deep fakes)","excerpt":"Deployers of an AI system that generates or manipulates image, audio or video content constituting a deep fake, shall disclose that the content has been artificially generated or manipulated.","isParaphrase":false},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"governs","citation":"§4.5 (content authentication, watermarking) — rescinded 20 Jan 2025 by EO 14148; successor EO 14179 is silent on deepfakes, leaving only NIST provenance artifacts","excerpt":null,"isParaphrase":false},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"governs","citation":"Art. 12 (labelling) + Deep Synthesis Rules","excerpt":null,"isParaphrase":false},{"shortCode":"G7-HIROSHIMA","name":"G7 Hiroshima AI Process Code of Conduct","slug":"g7-hiroshima-code","verdict":"governs","citation":"Code §5 (content provenance + watermarking)","excerpt":null,"isParaphrase":false},{"shortCode":"UN-RES-2024","name":"UN GA Resolution on Safe, Secure, Trustworthy AI","slug":"un-ai-resolution-2024","verdict":"implicit","citation":"References disinformation broadly","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF","name":"NIST AI Risk Management Framework","slug":"nist-ai-rmf","verdict":"implicit","citation":"GenAI Profile addresses synthetic content","excerpt":"Risks associated with transparency and accountability — as identified in the MAP function — are examined and documented.","isParaphrase":true},{"shortCode":"NIST-AI-RMF-GENAI","name":"NIST AI RMF Generative AI Profile","slug":"nist-ai-rmf-genai-profile","verdict":"governs","citation":"NIST AI 600-1 §3.11 Confabulation + §3.10 Information Integrity (synthetic content)","excerpt":null,"isParaphrase":false},{"shortCode":"IN-DPDP-2023","name":"India Digital Personal Data Protection Act + AI Advisory (MEITY)","slug":"india-dpdpa","verdict":"governs","citation":"MEITY Mar-2024 Advisory + IT Rules 2021 §3(1)(b)(v) deepfake takedown obligations","excerpt":null,"isParaphrase":false},{"shortCode":"WH-VOLUNTARY-2023","name":"White House Voluntary AI Commitments","slug":"wh-voluntary-2023","verdict":"governs","citation":"Commitments §5 (watermarking + content provenance for AI-generated content)","excerpt":null,"isParaphrase":false},{"shortCode":"SG-MODEL-AI-2024","name":"Singapore Model AI Governance Framework for Generative AI","slug":"singapore-model-ai-governance","verdict":"governs","citation":"Framework Dimension 7 — content provenance + synthetic-content disclosure","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-942","name":"California SB 942: AI Transparency Act","slug":"ca-sb-942","verdict":"implicit","citation":"'Deepfake' appears only in the SB 942 Legislative Counsel's Digest (a recital about a separate law), never in operative §§ 22757.1–22757.4; a deepfake produced by a covered provider's GenAI system is nonetheless a subset of the AI-generated image/video/audio reached by the § 22757.3(b) latent-disclosure and § 22757.2 detection duties","excerpt":null,"isParaphrase":false},{"shortCode":"CN-DEEPSYN-2022","name":"Provisions on the Administration of Deep Synthesis of Internet Information Services","slug":"china-deep-synthesis-provisions","verdict":"governs","citation":"Art. 17","excerpt":"深度合成服务提供者提供以下深度合成服务……应当在生成或者编辑的信息内容的合理位置、区域进行显著标识，向公众提示深度合成情况：……（三）人脸生成、人脸替换、人脸操控、姿态操控等人物图像、视频生成或者显著改变个人身份特征的编辑服务","isParaphrase":false},{"shortCode":"US-TAKEITDOWN-2025","name":"TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act)","slug":"take-it-down-act","verdict":"governs","citation":"Pub. L. 119-12 — criminalizes nonconsensual intimate 'digital forgeries' (AI deepfakes) of adults and minors and requires covered platforms to remove them within 48 hours; the statute names 'artificial intelligence' in its operative digital-forgery definition","excerpt":"'Digital forgery' [is an intimate visual depiction] created through the use of software, machine learning, artificial intelligence, or any other computer-generated or technological means … indistinguishable from an authentic visual depiction.","isParaphrase":true},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"governs","citation":"Art. 26(1)(c) inserts new Criminal Code Art. 612-quater: illicit dissemination of AI-generated or altered images/video/voices, without consent, apt to deceive and causing unjust harm — 1 to 5 years' imprisonment (querela-based; ex officio in aggravated cases).","excerpt":"«Art. 612-quater … Chiunque cagiona un danno ingiusto … diffondendo, senza il suo consenso, immagini, video o voci falsificati o alterati mediante l'impiego di sistemi di intelligenza artificiale … è punito con la reclusione da uno a cinque anni.»","isParaphrase":false}],"governsCount":11,"relatedConcepts":[{"code":"asl-3","label":"AI Safety Level 3 (ASL-3)","slug":"asl-3"},{"code":"red-team-evaluation","label":"Red-Team Evaluation","slug":"red-team-evaluation"},{"code":"provenance-watermarking","label":"Provenance & Watermarking","slug":"provenance-watermarking"}],"engagingArticles":["asl-3","ca-sb-942","china-deep-synthesis-provisions","china-genai-measures","deepfakes","eu-ai-act","g7-hiroshima-code","india-dpdpa","italy-ai-law-2025","nist-ai-rmf","nist-ai-rmf-genai-profile","provenance-watermarking","red-team-evaluation","singapore-model-ai-governance","take-it-down-act","un-ai-resolution-2024","us-eo-14110","wh-voluntary-2023"],"literatureCount":16,"convergence":[{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["ca-sb-942","china-genai-measures","deepfakes","g7-hiroshima-code","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":10},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["ca-sb-942","china-genai-measures","eu-ai-act","g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","un-ai-resolution-2024","us-eo-14110","wh-voluntary-2023"],"citeCount":10},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["asl-3","ca-sb-942","eu-ai-act","g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile","red-team-evaluation","singapore-model-ai-governance","us-eo-14110","wh-voluntary-2023"],"citeCount":10},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["ca-sb-942","china-deep-synthesis-provisions","deepfakes","g7-hiroshima-code","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","provenance-watermarking","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":10},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["ca-sb-942","deepfakes","g7-hiroshima-code","india-dpdpa","nist-ai-rmf-genai-profile","provenance-watermarking","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":8},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["asl-3","eu-ai-act","g7-hiroshima-code","nist-ai-rmf-genai-profile","red-team-evaluation","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":7},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["china-genai-measures","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":6},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile","un-ai-resolution-2024","us-eo-14110","wh-voluntary-2023"],"citeCount":6},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["ca-sb-942","china-genai-measures","nist-ai-rmf","singapore-model-ai-governance","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":6},{"key":"lit-groh-sankaranarayanan-singh-kim-lippman-picard-human-d","title":"Human detection of political speech deepfakes across transcripts, audio, and video","citingArticles":["ca-sb-942","china-deep-synthesis-provisions","provenance-watermarking","red-team-evaluation"],"citeCount":4},{"key":"lit-huijuan-peng-and-pey-woan-lee-reimagining-u-s-tort-law","title":"Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore","citingArticles":["nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":4},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["eu-ai-act","red-team-evaluation","us-eo-14110","wh-voluntary-2023"],"citeCount":4},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":3},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["g7-hiroshima-code","singapore-model-ai-governance","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":3},{"key":"lit-beatriz-kira-when-non-consensual-intimate-deepfakes-go","title":"When non-consensual intimate deepfakes go viral: The insufficiency of the UK Online Safety Act","citingArticles":["china-deep-synthesis-provisions","deepfakes","provenance-watermarking"],"citeCount":3},{"key":"lit-eloundou-manning-mishkin-rock-gpts-are-gpts-labor-mark","title":"GPTs are GPTs: Labor market impact potential of LLMs","citingArticles":["asl-3","ca-sb-942","red-team-evaluation"],"citeCount":3},{"key":"lit-keith-raymond-harris-ai-or-your-lying-eyes-some-shortc","title":"AI or Your Lying Eyes: Some Shortcomings of Artificially Intelligent Deepfake Detectors","citingArticles":["china-deep-synthesis-provisions","deepfakes","provenance-watermarking"],"citeCount":3},{"key":"lit-kyrie-zhixuan-zhou-abhinav-choudhry-ece-gumusel-and-ma","title":"'Sora is incredible and scary': public perceptions and governance challenges of text-to-video generative AI models","citingArticles":["china-genai-measures","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":3},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":3},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["china-genai-measures","india-dpdpa","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["india-dpdpa","nist-ai-rmf","singapore-model-ai-governance"],"citeCount":3},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":3},{"key":"lit-novelli-casolari-hacker-spedicato-floridi-generative-a","title":"Generative AI in EU law: Liability, privacy, intellectual property, and cybersecurity","citingArticles":["asl-3","china-deep-synthesis-provisions","provenance-watermarking"],"citeCount":3},{"key":"arxiv:2504.18236","title":null,"citingArticles":["nist-ai-rmf","singapore-model-ai-governance"],"citeCount":2},{"key":"lit-anderljung-barnhart-korinek-et-al-frontier-ai-regulati","title":"Frontier AI Regulation: Managing Emerging Risks to Public Safety","citingArticles":["asl-3","us-eo-14110"],"citeCount":2},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":2},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["india-dpdpa","un-ai-resolution-2024"],"citeCount":2}]},{"topicCode":"development_rights_framing","label":"Development-Rights Framings","url":"https://policywindow.org/wiki/synthesis/development-rights-framing","themeArticle":"https://policywindow.org/wiki/development-rights-framing","empiricalConsensus":"emerging","contestedQuestion":"Is development-rights framing compatible with the EU AIA's rights-based framing, or do they conflict on operational decisions (e.g., who can deploy frontier models in developing economies)?","stances":[{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"implicit","citation":"PRC has invoked development rights in UN AI debates (2024 GA)","excerpt":null,"isParaphrase":false},{"shortCode":"OECD-AI-PRIN","name":"OECD AI Principles (Recommendation)","slug":"oecd-ai-principles","verdict":"implicit","citation":"Principle 1.1 'inclusive growth' brushes against development-rights framing","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"implicit","citation":"Rights-based framing partly overlaps with development-rights doctrine but not explicitly","excerpt":null,"isParaphrase":false},{"shortCode":"UN-RES-2024","name":"UN GA Resolution on Safe, Secure, Trustworthy AI","slug":"un-ai-resolution-2024","verdict":"governs","citation":"Operative paragraphs frame AI through development-rights + digital divide lens; co-sponsored by Global-South coalition","excerpt":null,"isParaphrase":false},{"shortCode":"IN-DPDP-2023","name":"India Digital Personal Data Protection Act + AI Advisory (MEITY)","slug":"india-dpdpa","verdict":"governs","citation":"Digital India framing centres development rights + tech-sovereignty; explicit in DPDPA preamble + MEITY's AI Mission documents","excerpt":null,"isParaphrase":false},{"shortCode":"BR-AIBILL-2024","name":"Brazil AI Bill (PL 2338/2023)","slug":"brazil-ai-bill","verdict":"governs","citation":"PL 2338/2023 Arts. 3-4 (founding principles include 'sustainable development' + 'human dignity' — distinct from EU AIA's rights-only framing)","excerpt":null,"isParaphrase":false},{"shortCode":"ASEAN-AI-GUIDE-2024","name":"ASEAN Guide on AI Governance and Ethics","slug":"asean-ai-guide","verdict":"implicit","citation":"Guide centres 'pragmatic + flexible' implementation reflecting member-state development trajectories","excerpt":null,"isParaphrase":false},{"shortCode":"AU-AI-STRATEGY-2024","name":"African Union Continental AI Strategy","slug":"au-continental-ai-strategy","verdict":"governs","citation":"AU Strategy §§1-3 (AI as continental development priority + data-coloniality framing)","excerpt":null,"isParaphrase":false},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"governs","citation":"Policy Area 'Development and International Cooperation', para 79 (+ Diversity Principle para 67) — AI-for-development bound to the values/principles","excerpt":"Member States should ensure that the use of AI in areas of development such as education, science, culture... health care, agriculture... adheres to the values and principles set forth","isParaphrase":false},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"governs","citation":"Act No. 53 of 2025, Arts. 1 & 3(3)","excerpt":"... comprehensively and systematically advancing initiatives ... from basic research ... to their utilization in the daily lives of the public and in economic activities ...","isParaphrase":true},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"governs","citation":"GDC Objective 5, para 55(c) and capacity-building partnerships (A/RES/79/1, Annex I)","excerpt":"Help to build capacities, especially in developing countries, to access, develop, use and govern AI ... international partnerships on artificial intelligence capacity-building.","isParaphrase":true}],"governsCount":7,"relatedConcepts":[],"engagingArticles":["asean-ai-guide","au-continental-ai-strategy","brazil-ai-bill","china-genai-measures","coe-ai-convention","development-rights-framing","india-dpdpa","japan-ai-promotion-act","oecd-ai-principles","un-ai-resolution-2024","un-global-digital-compact","unesco-ai-ethics-recommendation"],"literatureCount":21,"convergence":[{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["asean-ai-guide","au-continental-ai-strategy","brazil-ai-bill","coe-ai-convention","development-rights-framing","india-dpdpa","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":9},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["asean-ai-guide","au-continental-ai-strategy","brazil-ai-bill","china-genai-measures","india-dpdpa","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":8},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["asean-ai-guide","au-continental-ai-strategy","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["brazil-ai-bill","coe-ai-convention","india-dpdpa","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["au-continental-ai-strategy","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["au-continental-ai-strategy","brazil-ai-bill","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["asean-ai-guide","china-genai-measures","india-dpdpa","oecd-ai-principles"],"citeCount":4},{"key":"arxiv:2504.18236","title":null,"citingArticles":["coe-ai-convention","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["asean-ai-guide","au-continental-ai-strategy","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["china-genai-measures","india-dpdpa","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["brazil-ai-bill","china-genai-measures","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["asean-ai-guide","brazil-ai-bill","oecd-ai-principles"],"citeCount":3},{"key":"lit-chiara-gallese-predictive-policing-and-predictive-just","title":"Predictive policing and predictive justice: Ethics, data protection, and the AI act","citingArticles":["coe-ai-convention","india-dpdpa"],"citeCount":2},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["asean-ai-guide","au-continental-ai-strategy"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["asean-ai-guide","au-continental-ai-strategy"],"citeCount":2},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["brazil-ai-bill","oecd-ai-principles"],"citeCount":2},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["china-genai-measures","un-ai-resolution-2024"],"citeCount":2},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["coe-ai-convention","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-pengfei-li-jianyi-yang-mohammad-a-islam-shaolei-ren-ma","title":"Making AI Less 'Thirsty': Uncovering and Addressing the Secret Water Footprint of AI Models","citingArticles":["asean-ai-guide","oecd-ai-principles"],"citeCount":2},{"key":"lit-stephen-weymouth-digital-disintegration-techno-blocs-a","title":"Digital Disintegration: Techno-Blocs and Strategic Sovereignty in the AI Era","citingArticles":["asean-ai-guide","au-continental-ai-strategy"],"citeCount":2},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["india-dpdpa","un-ai-resolution-2024"],"citeCount":2}]},{"topicCode":"environmental_impact_of_training","label":"Environmental Impact of AI Training","url":"https://policywindow.org/wiki/synthesis/environmental-impact-of-training","themeArticle":"https://policywindow.org/wiki/environmental-impact-of-training","empiricalConsensus":"emerging","contestedQuestion":"Should environmental obligations attach to (a) model-provider disclosure, (b) datacenter operator emissions caps, or (c) end-customer reporting? The training-vs-inference split also remains unresolved across instruments.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"implicit","citation":"Art. 95 voluntary codes of conduct include environmental sustainability; Recital 142 references energy efficiency reporting for GPAI","excerpt":"Codes of conduct may cover … assessing and minimising the impact of AI systems on environmental sustainability, including as regards energy-efficient programming and techniques for design…","isParaphrase":true},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"implicit","citation":"§5.2 directs environmental-review consideration; §4.2 reporting includes some energy data","excerpt":null,"isParaphrase":false},{"shortCode":"G7-HIROSHIMA","name":"G7 Hiroshima AI Process Code of Conduct","slug":"g7-hiroshima-code","verdict":"implicit","citation":"Code §6 references sustainable AI development; not detailed obligation","excerpt":null,"isParaphrase":false},{"shortCode":"OECD-AI-PRIN","name":"OECD AI Principles (Recommendation)","slug":"oecd-ai-principles","verdict":"implicit","citation":"Principle 1.1 inclusive growth + sustainable development; addresses environment implicitly","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"implicit","citation":"Art. 7 sustainability principle; environmental impact subsumed","excerpt":null,"isParaphrase":false},{"shortCode":"UN-RES-2024","name":"UN GA Resolution on Safe, Secure, Trustworthy AI","slug":"un-ai-resolution-2024","verdict":"implicit","citation":"Preamble references SDGs which include climate goals","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF-GENAI","name":"NIST AI RMF Generative AI Profile","slug":"nist-ai-rmf-genai-profile","verdict":"governs","citation":"NIST AI 600-1 — Environmental Impacts is one of 12 named GenAI risk categories","excerpt":null,"isParaphrase":false},{"shortCode":"ASEAN-AI-GUIDE-2024","name":"ASEAN Guide on AI Governance and Ethics","slug":"asean-ai-guide","verdict":"implicit","citation":"Guide references sustainable AI principles; not operationalised","excerpt":null,"isParaphrase":false},{"shortCode":"AU-AI-STRATEGY-2024","name":"African Union Continental AI Strategy","slug":"au-continental-ai-strategy","verdict":"implicit","citation":"Continental strategy includes sustainability themes; not operationalised","excerpt":null,"isParaphrase":false},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"governs","citation":"Policy Area 'Environment and Ecosystems', para 84 — assess direct/indirect environmental impact incl. carbon footprint + energy consumption","excerpt":"Member States and business enterprises should assess the direct and indirect environmental impact throughout the AI system life cycle, including... its carbon footprint, energy consumption","isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"implicit","citation":"Art. 3(1) lists 'sostenibilità' (sustainability) among the binding general principles governing AI development and use, alongside transparency, proportionality, security and non-discrimination. No operative environmental-reporting or training-footprint duty.","excerpt":"… nel rispetto … dei princìpi di trasparenza, proporzionalità, sicurezza, protezione dei dati personali, riservatezza, accuratezza, non discriminazione, parità dei sessi e sostenibilità.","isParaphrase":false},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"implicit","citation":"GDC para 11(e) lifecycle sustainability; Objective 5 narrative (A/RES/79/1, Annex I)","excerpt":"Promote sustainability across the life cycle of digital technologies ...; potential negative impacts of emerging digital technologies on ... the environment.","isParaphrase":true}],"governsCount":2,"relatedConcepts":[],"engagingArticles":["asean-ai-guide","au-continental-ai-strategy","coe-ai-convention","environmental-impact-of-training","eu-ai-act","g7-hiroshima-code","italy-ai-law-2025","nist-ai-rmf-genai-profile","oecd-ai-principles","un-ai-resolution-2024","un-global-digital-compact","unesco-ai-ethics-recommendation","us-eo-14110"],"literatureCount":13,"convergence":[{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["asean-ai-guide","au-continental-ai-strategy","environmental-impact-of-training","g7-hiroshima-code","nist-ai-rmf-genai-profile","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":8},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["asean-ai-guide","eu-ai-act","g7-hiroshima-code","nist-ai-rmf-genai-profile","oecd-ai-principles","us-eo-14110"],"citeCount":6},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["asean-ai-guide","au-continental-ai-strategy","coe-ai-convention","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":6},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["eu-ai-act","g7-hiroshima-code","nist-ai-rmf-genai-profile","un-ai-resolution-2024","us-eo-14110"],"citeCount":5},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["asean-ai-guide","au-continental-ai-strategy","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["asean-ai-guide","au-continental-ai-strategy","g7-hiroshima-code","un-ai-resolution-2024"],"citeCount":4},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["au-continental-ai-strategy","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile","un-ai-resolution-2024","us-eo-14110"],"citeCount":4},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["eu-ai-act","g7-hiroshima-code","nist-ai-rmf-genai-profile","oecd-ai-principles"],"citeCount":4},{"key":"arxiv:2504.18236","title":null,"citingArticles":["coe-ai-convention","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["au-continental-ai-strategy","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["asean-ai-guide","nist-ai-rmf-genai-profile","oecd-ai-principles"],"citeCount":3},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["coe-ai-convention","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-pengfei-li-jianyi-yang-mohammad-a-islam-shaolei-ren-ma","title":"Making AI Less 'Thirsty': Uncovering and Addressing the Secret Water Footprint of AI Models","citingArticles":["asean-ai-guide","environmental-impact-of-training","oecd-ai-principles"],"citeCount":3},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile"],"citeCount":2},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["asean-ai-guide","au-continental-ai-strategy"],"citeCount":2},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["asean-ai-guide","au-continental-ai-strategy"],"citeCount":2},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile"],"citeCount":2},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile"],"citeCount":2},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["coe-ai-convention","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-stephen-weymouth-digital-disintegration-techno-blocs-a","title":"Digital Disintegration: Techno-Blocs and Strategic Sovereignty in the AI Era","citingArticles":["asean-ai-guide","au-continental-ai-strategy"],"citeCount":2}]},{"topicCode":"foundation_models","label":"Foundation Models / GPAI","url":"https://policywindow.org/wiki/synthesis/foundation-models","themeArticle":"https://policywindow.org/wiki/foundation-models","empiricalConsensus":"contested","contestedQuestion":"Does the foundation-model category map to a coherent capability tier, or is it a regulatory convenience? Compute-threshold vs behavioural-threshold debate is unresolved across EU/US/China.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Arts. 51-55 (general-purpose AI + systemic risk)","excerpt":"A general-purpose AI model shall be classified as a general-purpose AI model with systemic risk if it meets any of the following conditions: (a) it has high impact capabilities…","isParaphrase":false},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"governs","citation":"§4.2(a) — Defense Production Act reporting","excerpt":null,"isParaphrase":false},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"Cross-cutting principles; sector regulators apply","excerpt":null,"isParaphrase":false},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"governs","citation":"Art. 2 (applies to GenAI services regardless of size)","excerpt":null,"isParaphrase":false},{"shortCode":"G7-HIROSHIMA","name":"G7 Hiroshima AI Process Code of Conduct","slug":"g7-hiroshima-code","verdict":"governs","citation":"Code applies to advanced AI","excerpt":null,"isParaphrase":false},{"shortCode":"OECD-AI-PRIN","name":"OECD AI Principles (Recommendation)","slug":"oecd-ai-principles","verdict":"implicit","citation":"2024 update clarifies GPAI scope","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"implicit","citation":"Applies to AI throughout lifecycle (Art. 3)","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF","name":"NIST AI Risk Management Framework","slug":"nist-ai-rmf","verdict":"governs","citation":"GenAI Profile (NIST AI 600-1, 2024)","excerpt":"Intended purposes, potentially beneficial uses, context-specific laws, norms and expectations, and prospective settings in which the AI system will be deployed are understood and documented.","isParaphrase":false},{"shortCode":"BLETCHLEY-2023","name":"Bletchley Declaration on AI Safety","slug":"bletchley-declaration","verdict":"governs","citation":"Declaration §1-2 (frontier AI defined as the subject)","excerpt":null,"isParaphrase":false},{"shortCode":"SEOUL-2024","name":"Seoul Declaration on Safe, Innovative and Inclusive AI","slug":"seoul-declaration","verdict":"governs","citation":"Declaration + accompanying Frontier AI Safety Commitments (16 signatory companies)","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF-GENAI","name":"NIST AI RMF Generative AI Profile","slug":"nist-ai-rmf-genai-profile","verdict":"governs","citation":"Entire NIST AI 600-1 scope is GPAI / GenAI","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-1047","name":"California SB-1047: Safe and Secure Innovation for Frontier AI Models Act","slug":"ca-sb-1047","verdict":"governs","citation":"Cal. SB-1047 §22602 — 'covered model' = trained with >10^26 operations AND >$100M cost (or fine-tuning >$10M); vetoed 29 Sep 2024","excerpt":null,"isParaphrase":false},{"shortCode":"IN-DPDP-2023","name":"India Digital Personal Data Protection Act + AI Advisory (MEITY)","slug":"india-dpdpa","verdict":"implicit","citation":"MEITY Apr-2024 advisory walked back the Mar-2024 pre-deployment-approval requirement; current approach is post-deployment incident reporting","excerpt":null,"isParaphrase":false},{"shortCode":"BR-AIBILL-2024","name":"Brazil AI Bill (PL 2338/2023)","slug":"brazil-ai-bill","verdict":"governs","citation":"PL 2338/2023 Arts. 17-19 (general-purpose AI systemic-risk obligations)","excerpt":null,"isParaphrase":false},{"shortCode":"ASEAN-AI-GUIDE-2024","name":"ASEAN Guide on AI Governance and Ethics","slug":"asean-ai-guide","verdict":"implicit","citation":"Guide §6 covers GenAI but with flexible implementation expectations","excerpt":null,"isParaphrase":false},{"shortCode":"ANTHROPIC-RSP-2024","name":"Anthropic Responsible Scaling Policy (RSP) v2","slug":"anthropic-rsp","verdict":"governs","citation":"RSP v2 §2 — ASL framework applies to frontier model releases","excerpt":null,"isParaphrase":false},{"shortCode":"OPENAI-PREPAREDNESS-2023","name":"OpenAI Preparedness Framework","slug":"openai-preparedness","verdict":"governs","citation":"Preparedness Framework §1-2 — applies to all OpenAI frontier-model releases","excerpt":null,"isParaphrase":false},{"shortCode":"DEEPMIND-FSF-2024","name":"Google DeepMind Frontier Safety Framework","slug":"deepmind-fsf","verdict":"governs","citation":"FSF applies to Google DeepMind frontier-model releases","excerpt":null,"isParaphrase":false},{"shortCode":"META-FRONTIER-2024","name":"Meta Frontier AI Framework","slug":"meta-frontier-ai-framework","verdict":"governs","citation":"Framework applies to Meta frontier-model releases (Llama family)","excerpt":null,"isParaphrase":false},{"shortCode":"UK-US-AISI-MOU-2024","name":"UK-US AI Safety Institute Memorandum of Understanding","slug":"uk-us-aisi-mou","verdict":"governs","citation":"MoU scope is frontier AI evaluation","excerpt":null,"isParaphrase":false},{"shortCode":"WH-VOLUNTARY-2023","name":"White House Voluntary AI Commitments","slug":"wh-voluntary-2023","verdict":"governs","citation":"Commitments §1-2 — internal + external security testing of frontier models","excerpt":null,"isParaphrase":false},{"shortCode":"SG-MODEL-AI-2024","name":"Singapore Model AI Governance Framework for Generative AI","slug":"singapore-model-ai-governance","verdict":"governs","citation":"Framework Dimension 3 (Trusted Development + Deployment) explicitly covers GenAI models","excerpt":null,"isParaphrase":false},{"shortCode":"JP-METI-AI-2024","name":"Japan METI AI Guidelines for Business","slug":"japan-meti-ai-guidelines","verdict":"governs","citation":"Guidelines Part 3 — covers AI providers including foundation-model developers","excerpt":null,"isParaphrase":false},{"shortCode":"EU-GPAI-COP-2025","name":"EU General-Purpose AI Code of Practice","slug":"gpai-code-of-practice","verdict":"governs","citation":"Chapter 3 (Safety & Security) operationalises Art. 55 systemic-risk-tier obligations for GPAI providers","excerpt":null,"isParaphrase":false},{"shortCode":"OMB-M-24-10","name":"OMB Memorandum M-24-10 (Advancing Governance, Innovation, and Risk Management for Agency Use of AI)","slug":"omb-m-24-10","verdict":"implicit","citation":"§5 + Attachment 1 — minimum practices apply to safety- + rights-impacting AI regardless of foundation-model classification; no compute-threshold trigger","excerpt":"Before agencies use new or existing safety-impacting or rights-impacting AI, they must implement the minimum practices in this section; if they cannot, they must cease using the AI until compliance is achieved.","isParaphrase":true},{"shortCode":"GSA-AI-GUIDE-2024","name":"GSA Generative AI and Specialized Computing Infrastructure Acquisition Resource Guide","slug":"gsa-ai-acquisition-guide","verdict":"governs","citation":"Sections posing generative-AI vendor-evaluation + model-provenance due-diligence questions for contracting officers","excerpt":"Faithful summary: the guide treats generative-AI and foundation-model acquisition as a discrete category, posing due-diligence questions for evaluating model provenance, capabilities, and vendor documentation.","isParaphrase":true},{"shortCode":"DOD-RAI-2022","name":"DoD Responsible AI Strategy and Implementation Pathway","slug":"dod-rai-strategy","verdict":"implicit","citation":"Tenet 3 (AI Product and Acquisition Lifecycle) + Tenet 5 (Responsible AI Ecosystem) — RAI integration applies regardless of model architecture; foundation-model-specific obligations flow through CDAO RAI Toolkit guidance","excerpt":null,"isParaphrase":false},{"shortCode":"FEDRAMP-AI-2024","name":"FedRAMP AI Cloud Procurement Guidance","slug":"fedramp-ai-guidance","verdict":"implicit","citation":"GenAI-specific control tailoring guidance addresses model-specific risks (training-data exposure, prompt-injection, output disclosure) within SSP + NIST SP 800-53 control overlay selection","excerpt":null,"isParaphrase":false},{"shortCode":"DFARS-252-204","name":"DFARS Subpart 252.204 (Safeguarding Covered Defense Information and Cyber Incident Reporting)","slug":"dfars-252-204","verdict":"implicit","citation":"252.204-7012 — AI-system source code, model weights, training data fall within Covered Defense Information scope when the underlying contract designates these as CDI; foundation-model artefacts are CDI through the standard contract designation pathway","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-53","name":"California SB-53: Transparency in Frontier Artificial Intelligence Act (TFAIA)","slug":"ca-sb-53","verdict":"governs","citation":"Bus. & Prof. Code § 22757.11 — defines 'foundation model' + 'frontier model' (>10^26 FLOP) as the regulated class","excerpt":"'Frontier model' means a foundation model that was trained using a quantity of computing power greater than 10^26 integer or floating-point operations, including the computing power used in subsequent fine-tuning or modifications.","isParaphrase":false},{"shortCode":"CA-SB-942","name":"California SB 942: AI Transparency Act","slug":"ca-sb-942","verdict":"implicit","citation":"No operative provision regulates foundation models as a class; the regulated party ('covered provider', § 22757.1) is defined by an output/scale hook — a producer of a publicly-accessible GenAI system with over 1,000,000 monthly users — so a foundation-model producer is reached only incidentally via the § 22757.2–.3 output-disclosure duties, not by any model-level obligation","excerpt":null,"isParaphrase":false},{"shortCode":"NY-RAISE-2025","name":"New York RAISE Act: Responsible AI Safety and Education Act","slug":"ny-raise-act","verdict":"governs","citation":"N.Y. Gen. Bus. Law § 1420(6) defines 'frontier model' (>10^26 FLOP, >$100M compute) + § 1421 imposes operative pre-deployment duties on large frontier-model developers","excerpt":"'Frontier model' means an AI model trained using greater than 10^26 computational operations, the compute cost of which exceeds one hundred million dollars (or a model knowledge-distilled from such a model).","isParaphrase":true},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"implicit","citation":"Act No. 53 of 2025, Arts. 2 & 12","excerpt":"\"AI-related technology\" means technology ... that ... substitute[s] for intellectual abilities involved in human cognition, reasoning, and judgment ...","isParaphrase":true},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"implicit","citation":"GDC Objective 5 (A/RES/79/1, Annex I)","excerpt":"open artificial intelligence models and systems ... evidence-based impact, risk and opportunity assessments.","isParaphrase":false}],"governsCount":22,"relatedConcepts":[{"code":"frontier-tier","label":"Frontier-Tier AI","slug":"frontier-tier"},{"code":"asl-3","label":"AI Safety Level 3 (ASL-3)","slug":"asl-3"},{"code":"systemic-risk","label":"Systemic Risk (AI)","slug":"systemic-risk"},{"code":"designated-systemic","label":"Designated Systemic-Risk Model","slug":"designated-systemic"},{"code":"compute-threshold","label":"Compute Threshold (AI Governance)","slug":"compute-threshold"},{"code":"red-team-evaluation","label":"Red-Team Evaluation","slug":"red-team-evaluation"},{"code":"model-card","label":"Model Card","slug":"model-card"},{"code":"alignment","label":"AI Alignment","slug":"alignment"},{"code":"deceptive-alignment","label":"Deceptive Alignment","slug":"deceptive-alignment"},{"code":"mesa-optimization","label":"Mesa-Optimization","slug":"mesa-optimization"},{"code":"scalable-oversight","label":"Scalable Oversight","slug":"scalable-oversight"},{"code":"capability-elicitation","label":"Capability Elicitation","slug":"capability-elicitation"},{"code":"dual-use-research-taxonomy","label":"Dual-Use Research Norms (DURC for AI)","slug":"dual-use-research-taxonomy"},{"code":"policy-instrument","label":"Policy Instrument","slug":"policy-instrument"},{"code":"ai-supply-chain","label":"AI Supply Chain","slug":"ai-supply-chain"},{"code":"prompt-injection","label":"Prompt Injection","slug":"prompt-injection"},{"code":"agentic-system","label":"Agentic AI System","slug":"agentic-system"},{"code":"tool-use-safety","label":"Tool-Use Safety","slug":"tool-use-safety"},{"code":"multi-turn-evaluation","label":"Multi-Turn Evaluation","slug":"multi-turn-evaluation"},{"code":"data-poisoning","label":"Data Poisoning","slug":"data-poisoning"},{"code":"model-distillation-risk","label":"Model Distillation Risk","slug":"model-distillation-risk"},{"code":"jailbreak-resistance","label":"Jailbreak Resistance","slug":"jailbreak-resistance"},{"code":"model-merging-risk","label":"Model-Merging Risk","slug":"model-merging-risk"},{"code":"inference-time-compute","label":"Inference-Time Compute","slug":"inference-time-compute"},{"code":"sandbagging","label":"Sandbagging","slug":"sandbagging"},{"code":"hallucination","label":"Hallucination","slug":"hallucination"},{"code":"in-context-learning","label":"In-Context Learning","slug":"in-context-learning"},{"code":"retrieval-augmented-generation","label":"Retrieval-Augmented Generation (RAG)","slug":"retrieval-augmented-generation"},{"code":"chain-of-thought-monitoring","label":"Chain-of-Thought Monitoring","slug":"chain-of-thought-monitoring"}],"engagingArticles":["agentic-system","ai-supply-chain","alignment","anthropic-rsp","asean-ai-guide","asl-3","bletchley-declaration","brazil-ai-bill","ca-sb-1047","ca-sb-53","ca-sb-942","capability-elicitation","chain-of-thought-monitoring","china-genai-measures","coe-ai-convention","compute-threshold","data-poisoning","deceptive-alignment","deepmind-fsf","designated-systemic","dfars-252-204","dod-rai-strategy","dual-use-research-taxonomy","eu-ai-act","fedramp-ai-guidance","foundation-models","frontier-tier","g7-hiroshima-code","gpai-code-of-practice","gsa-ai-acquisition-guide","hallucination","in-context-learning","india-dpdpa","inference-time-compute","jailbreak-resistance","japan-ai-promotion-act","japan-meti-ai-guidelines","mesa-optimization","meta-frontier-ai-framework","model-card","model-distillation-risk","model-merging-risk","multi-turn-evaluation","nist-ai-rmf","nist-ai-rmf-genai-profile","ny-raise-act","oecd-ai-principles","omb-m-24-10","openai-preparedness","policy-instrument","prompt-injection","red-team-evaluation","retrieval-augmented-generation","sandbagging","scalable-oversight","seoul-declaration","singapore-model-ai-governance","systemic-risk","tool-use-safety","uk-ai-white-paper","uk-us-aisi-mou","un-global-digital-compact","us-eo-14110","wh-voluntary-2023"],"literatureCount":40,"convergence":[{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["ai-supply-chain","anthropic-rsp","asean-ai-guide","asl-3","bletchley-declaration","brazil-ai-bill","ca-sb-1047","ca-sb-53","ca-sb-942","capability-elicitation","data-poisoning","deepmind-fsf","designated-systemic","dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","foundation-models","frontier-tier","g7-hiroshima-code","gpai-code-of-practice","gsa-ai-acquisition-guide","in-context-learning","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-card","model-distillation-risk","model-merging-risk","multi-turn-evaluation","nist-ai-rmf","nist-ai-rmf-genai-profile","oecd-ai-principles","openai-preparedness","policy-instrument","red-team-evaluation","retrieval-augmented-generation","seoul-declaration","singapore-model-ai-governance","systemic-risk","tool-use-safety","uk-ai-white-paper","uk-us-aisi-mou","us-eo-14110","wh-voluntary-2023"],"citeCount":44},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["anthropic-rsp","asl-3","bletchley-declaration","brazil-ai-bill","ca-sb-1047","ca-sb-53","capability-elicitation","data-poisoning","deepmind-fsf","designated-systemic","dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","foundation-models","g7-hiroshima-code","gpai-code-of-practice","gsa-ai-acquisition-guide","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-card","model-merging-risk","multi-turn-evaluation","nist-ai-rmf-genai-profile","oecd-ai-principles","openai-preparedness","red-team-evaluation","seoul-declaration","singapore-model-ai-governance","systemic-risk","tool-use-safety","uk-ai-white-paper","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":34},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["ai-supply-chain","anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","capability-elicitation","compute-threshold","designated-systemic","dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","frontier-tier","gsa-ai-acquisition-guide","in-context-learning","inference-time-compute","model-distillation-risk","omb-m-24-10","openai-preparedness","policy-instrument","red-team-evaluation","seoul-declaration","systemic-risk","us-eo-14110","wh-voluntary-2023"],"citeCount":25},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["ai-supply-chain","asean-ai-guide","bletchley-declaration","ca-sb-1047","china-genai-measures","data-poisoning","dfars-252-204","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","india-dpdpa","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-card","nist-ai-rmf","nist-ai-rmf-genai-profile","oecd-ai-principles","retrieval-augmented-generation","singapore-model-ai-governance","uk-ai-white-paper","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":22},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","deepmind-fsf","g7-hiroshima-code","gpai-code-of-practice","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-ai-white-paper","uk-us-aisi-mou","us-eo-14110","wh-voluntary-2023"],"citeCount":16},{"key":"lit-mary-phuong-matthew-aitchison-elliot-catt-et-al-google","title":"Evaluating Frontier Models for Dangerous Capabilities","citingArticles":["asl-3","capability-elicitation","designated-systemic","dual-use-research-taxonomy","frontier-tier","in-context-learning","inference-time-compute","jailbreak-resistance","mesa-optimization","model-card","model-distillation-risk","model-merging-risk","multi-turn-evaluation","systemic-risk","tool-use-safety"],"citeCount":15},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["brazil-ai-bill","ca-sb-1047","coe-ai-convention","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","hallucination","india-dpdpa","japan-meti-ai-guidelines","model-card","nist-ai-rmf","oecd-ai-principles","omb-m-24-10","singapore-model-ai-governance","uk-ai-white-paper"],"citeCount":15},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["ai-supply-chain","capability-elicitation","compute-threshold","designated-systemic","dfars-252-204","fedramp-ai-guidance","frontier-tier","gsa-ai-acquisition-guide","in-context-learning","inference-time-compute","model-distillation-risk","policy-instrument","systemic-risk","us-eo-14110"],"citeCount":14},{"key":"lit-sastry-heim-belfield-anderljung-brundage-et-al-computi","title":"Computing Power and the Governance of Artificial Intelligence","citingArticles":["ai-supply-chain","ca-sb-1047","capability-elicitation","compute-threshold","designated-systemic","dfars-252-204","frontier-tier","gsa-ai-acquisition-guide","inference-time-compute","model-distillation-risk","omb-m-24-10","policy-instrument","red-team-evaluation","systemic-risk"],"citeCount":14},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","deepmind-fsf","dod-rai-strategy","gpai-code-of-practice","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","tool-use-safety","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":13},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["anthropic-rsp","brazil-ai-bill","ca-sb-942","china-genai-measures","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","japan-meti-ai-guidelines","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","us-eo-14110","wh-voluntary-2023"],"citeCount":13},{"key":"lit-eloundou-manning-mishkin-rock-gpts-are-gpts-labor-mark","title":"GPTs are GPTs: Labor market impact potential of LLMs","citingArticles":["asl-3","ca-sb-942","capability-elicitation","dual-use-research-taxonomy","fedramp-ai-guidance","foundation-models","gsa-ai-acquisition-guide","in-context-learning","model-distillation-risk","omb-m-24-10","policy-instrument","red-team-evaluation","retrieval-augmented-generation"],"citeCount":13},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":12},{"key":"arxiv:2504.18236","title":null,"citingArticles":["ca-sb-1047","coe-ai-convention","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","hallucination","japan-meti-ai-guidelines","nist-ai-rmf","oecd-ai-principles","singapore-model-ai-governance","uk-ai-white-paper"],"citeCount":11},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","uk-us-aisi-mou"],"citeCount":11},{"key":"lit-novelli-casolari-hacker-spedicato-floridi-generative-a","title":"Generative AI in EU law: Liability, privacy, intellectual property, and cybersecurity","citingArticles":["ai-supply-chain","asl-3","data-poisoning","dfars-252-204","dual-use-research-taxonomy","foundation-models","hallucination","japan-meti-ai-guidelines","model-card","model-merging-risk","retrieval-augmented-generation"],"citeCount":11},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["anthropic-rsp","ca-sb-53","deepmind-fsf","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":10},{"key":"lit-anderljung-barnhart-korinek-et-al-frontier-ai-regulati","title":"Frontier AI Regulation: Managing Emerging Risks to Public Safety","citingArticles":["asl-3","capability-elicitation","dual-use-research-taxonomy","foundation-models","frontier-tier","model-card","model-merging-risk","multi-turn-evaluation","policy-instrument","us-eo-14110"],"citeCount":10},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["ai-supply-chain","data-poisoning","gpai-code-of-practice","gsa-ai-acquisition-guide","india-dpdpa","meta-frontier-ai-framework","model-merging-risk","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":9},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["bletchley-declaration","ca-sb-1047","deepmind-fsf","dod-rai-strategy","gpai-code-of-practice","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":9},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["ca-sb-942","china-genai-measures","g7-hiroshima-code","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":8},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["ca-sb-942","g7-hiroshima-code","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":7},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["asean-ai-guide","japan-meti-ai-guidelines","openai-preparedness","policy-instrument","singapore-model-ai-governance","uk-us-aisi-mou"],"citeCount":6},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["ca-sb-942","china-genai-measures","japan-meti-ai-guidelines","nist-ai-rmf","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":6},{"key":"lit-akash-r-wasil-tom-reed-jack-william-miller-peter-barne","title":"Verification methods for international AI agreements","citingArticles":["capability-elicitation","compute-threshold","designated-systemic","inference-time-compute","model-distillation-risk"],"citeCount":5},{"key":"lit-hacker-engel-mauer-regulating-chatgpt-and-other-large","title":"Regulating ChatGPT and other Large Generative AI Models","citingArticles":["asl-3","foundation-models","model-card","model-merging-risk","multi-turn-evaluation"],"citeCount":5},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["asean-ai-guide","brazil-ai-bill","coe-ai-convention","india-dpdpa","oecd-ai-principles"],"citeCount":5},{"key":"lit-lehdonvirta-w-hawkins-compute-north-vs-compute-south-t","title":"Compute North vs. Compute South: The Uneven Possibilities of Compute-based AI Governance Around the Globe","citingArticles":["ai-supply-chain","compute-threshold","designated-systemic","frontier-tier","policy-instrument"],"citeCount":5},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["asean-ai-guide","brazil-ai-bill","china-genai-measures","india-dpdpa","oecd-ai-principles"],"citeCount":5},{"key":"lit-maksym-andriushchenko-alexandra-souly-mateusz-dziemian","title":"AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents","citingArticles":["bletchley-declaration","deepmind-fsf","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":5},{"key":"lit-sarah-sterz-kevin-baum-sebastian-biewer-holger-hermann","title":"On the Quest for Effectiveness in Human Oversight: Interdisciplinary Perspectives","citingArticles":["data-poisoning","hallucination","model-card","multi-turn-evaluation","retrieval-augmented-generation"],"citeCount":5},{"key":"lit-stephen-weymouth-digital-disintegration-techno-blocs-a","title":"Digital Disintegration: Techno-Blocs and Strategic Sovereignty in the AI Era","citingArticles":["ai-supply-chain","asean-ai-guide","compute-threshold","model-distillation-risk","policy-instrument"],"citeCount":5},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["ca-sb-942","g7-hiroshima-code","india-dpdpa","nist-ai-rmf-genai-profile","wh-voluntary-2023"],"citeCount":5},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["asean-ai-guide","g7-hiroshima-code","nist-ai-rmf-genai-profile","oecd-ai-principles"],"citeCount":4},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["asean-ai-guide","g7-hiroshima-code","singapore-model-ai-governance","uk-ai-white-paper"],"citeCount":4},{"key":"lit-anka-reuel-ben-bucknall-stephen-casper-tim-fist-lennar","title":"Open Problems in Technical AI Governance","citingArticles":["capability-elicitation","compute-threshold","in-context-learning","multi-turn-evaluation"],"citeCount":4},{"key":"lit-bengio-hinton-yao-song-et-al-managing-extreme-ai-risks","title":"Managing extreme AI risks amid rapid progress","citingArticles":["ca-sb-1047","dod-rai-strategy","tool-use-safety","uk-us-aisi-mou"],"citeCount":4},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["asean-ai-guide","japan-meti-ai-guidelines","openai-preparedness","policy-instrument"],"citeCount":4},{"key":"lit-david-m-allison-stephen-herzog-artificial-intelligence","title":"Artificial Intelligence and Nuclear Weapons Proliferation: The Technological Arms Race for (In)visibility","citingArticles":["bletchley-declaration","dod-rai-strategy","meta-frontier-ai-framework","openai-preparedness"],"citeCount":4},{"key":"lit-francesca-palmiotto-the-ai-act-roller-coaster-the-evol","title":"The AI Act Roller Coaster: The Evolution of Fundamental Rights Protection in the Legislative Process and the Future of the Regulation","citingArticles":["dfars-252-204","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":4},{"key":"lit-huijuan-peng-and-pey-woan-lee-reimagining-u-s-tort-law","title":"Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore","citingArticles":["nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":4},{"key":"lit-kyrie-zhixuan-zhou-abhinav-choudhry-ece-gumusel-and-ma","title":"'Sora is incredible and scary': public perceptions and governance challenges of text-to-video generative AI models","citingArticles":["china-genai-measures","japan-meti-ai-guidelines","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":4},{"key":"lit-lennart-heim-tim-fist-janet-egan-sihao-huang-stephen-z","title":"Governing Through the Cloud: The Intermediary Role of Compute Providers in AI Regulation","citingArticles":["compute-threshold","designated-systemic","inference-time-compute","model-distillation-risk"],"citeCount":4},{"key":"lit-martin-kretschmer-bartolomeo-meletti-lionel-bently-gab","title":"Copyright and AI in the UK: Opting-In or Opting-Out?","citingArticles":["ai-supply-chain","gpai-code-of-practice","gsa-ai-acquisition-guide","retrieval-augmented-generation"],"citeCount":4},{"key":"arxiv:2407.05694","title":null,"citingArticles":["compute-threshold","foundation-models","systemic-risk"],"citeCount":3},{"key":"concept-mesa-optimization","title":"Mesa-Optimization","citingArticles":["alignment","deceptive-alignment","mesa-optimization"],"citeCount":3},{"key":"lit-chiara-gallese-predictive-policing-and-predictive-just","title":"Predictive policing and predictive justice: Ethics, data protection, and the AI act","citingArticles":["coe-ai-convention","india-dpdpa","uk-ai-white-paper"],"citeCount":3},{"key":"lit-ezgi-yazici-toward-a-global-standard-for-ethical-ai-re","title":"Toward a global standard for ethical AI regulation: addressing gaps in AI-driven biometric and high-resolution satellite imaging in the EU AI Act","citingArticles":["dfars-252-204","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":3},{"key":"lit-gary-e-weissman-toni-mankowitz-genevieve-p-kanter-unre","title":"Unregulated large language models produce medical device-like output","citingArticles":["hallucination","omb-m-24-10","uk-ai-white-paper"],"citeCount":3},{"key":"lit-henrik-palmer-olsen-thomas-troels-hildebrandt-corneliu","title":"The Right to Transparency in Public Governance: Freedom of Information and the Use of Artificial Intelligence by Public Agencies","citingArticles":["in-context-learning","model-card","multi-turn-evaluation"],"citeCount":3},{"key":"lit-justin-kollar-andrew-stokols-geopolitical-ecologies-of","title":"Geopolitical ecologies of cloud capitalism: Territorial restructuring and the making of national computing power in the U.S. and China","citingArticles":["ai-supply-chain","ca-sb-53","compute-threshold"],"citeCount":3},{"key":"lit-longpre-mahari-et-al-data-provenance-initiative-a-larg","title":"A large-scale audit of dataset licensing and attribution in AI","citingArticles":["data-poisoning","model-merging-risk","retrieval-augmented-generation"],"citeCount":3},{"key":"lit-megha-shrivastava-and-amrita-jash-china-s-semiconducto","title":"China's semiconductor conundrum: understanding US export controls and their efficacy","citingArticles":["anthropic-rsp","meta-frontier-ai-framework","us-eo-14110"],"citeCount":3},{"key":"lit-petros-terzis-michael-veale-no-lle-gaumann-law-and-the","title":"Law and the Emerging Political Economy of Algorithmic Audits","citingArticles":["data-poisoning","model-card","multi-turn-evaluation"],"citeCount":3},{"key":"lit-shayne-longpre-robert-mahari-ariel-lee-et-al-consent-i","title":"Consent in Crisis: The Rapid Decline of the AI Data Commons","citingArticles":["data-poisoning","dual-use-research-taxonomy","model-merging-risk"],"citeCount":3},{"key":"arxiv:1606.06565","title":null,"citingArticles":["alignment","scalable-oversight"],"citeCount":2},{"key":"arxiv:1706.03741","title":null,"citingArticles":["alignment","scalable-oversight"],"citeCount":2},{"key":"arxiv:2311.08379","title":null,"citingArticles":["deceptive-alignment","sandbagging"],"citeCount":2},{"key":"arxiv:2404.13208","title":null,"citingArticles":["prompt-injection","tool-use-safety"],"citeCount":2},{"key":"arxiv:2412.14093","title":null,"citingArticles":["deceptive-alignment","sandbagging"],"citeCount":2},{"key":"concept-prompt-injection","title":"Prompt Injection","citingArticles":["prompt-injection","retrieval-augmented-generation"],"citeCount":2},{"key":"lit-aditya-loganathan-michael-friedman-tayab-waseem-et-al","title":"Current state of Food and Drug Administration-approved artificial intelligence/machine learning medical devices: pathways, transparency, and evidence gaps","citingArticles":["hallucination","uk-ai-white-paper"],"citeCount":2},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["japan-meti-ai-guidelines","oecd-ai-principles"],"citeCount":2},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["brazil-ai-bill","japan-meti-ai-guidelines"],"citeCount":2},{"key":"lit-groh-sankaranarayanan-singh-kim-lippman-picard-human-d","title":"Human detection of political speech deepfakes across transcripts, audio, and video","citingArticles":["ca-sb-942","red-team-evaluation"],"citeCount":2},{"key":"lit-gutierrez-aguirre-uuk-boine-franklin-a-proposal-for-a","title":"A Proposal for a Definition of General Purpose Artificial Intelligence Systems","citingArticles":["frontier-tier","model-card"],"citeCount":2},{"key":"lit-jared-kaplan-sam-mccandlish-tom-henighan-tom-b-brown-b","title":"Scaling Laws for Neural Language Models","citingArticles":["compute-threshold","foundation-models"],"citeCount":2},{"key":"lit-jonas-schuett-markus-anderljung-alexis-carlier-leonie","title":"From Principles to Rules: A Regulatory Approach for Frontier AI","citingArticles":["jailbreak-resistance","tool-use-safety"],"citeCount":2},{"key":"lit-jordan-hoffmann-sebastian-borgeaud-arthur-mensch-et-al","title":"Training Compute-Optimal Large Language Models","citingArticles":["compute-threshold","foundation-models"],"citeCount":2},{"key":"lit-natalie-sheard-algorithm-facilitated-discrimination-a","title":"Algorithm-facilitated discrimination: a socio-legal study of the use by employers of artificial intelligence hiring systems","citingArticles":["omb-m-24-10","uk-ai-white-paper"],"citeCount":2},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["coe-ai-convention","uk-ai-white-paper"],"citeCount":2},{"key":"lit-pengfei-li-jianyi-yang-mohammad-a-islam-shaolei-ren-ma","title":"Making AI Less 'Thirsty': Uncovering and Addressing the Secret Water Footprint of AI Models","citingArticles":["asean-ai-guide","oecd-ai-principles"],"citeCount":2}]},{"topicCode":"redress","label":"Individual Redress","url":"https://policywindow.org/wiki/synthesis/redress","themeArticle":"https://policywindow.org/wiki/redress","empiricalConsensus":"settled","contestedQuestion":null,"stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Art. 85 (right to lodge complaints)","excerpt":"any natural or legal person … may lodge a complaint with the relevant market surveillance authority … [where] there are grounds to consider that there has been an infringement of this Regulation.","isParaphrase":true},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"Principle 5 (contestability + redress)","excerpt":null,"isParaphrase":false},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"governs","citation":"Art. 15 (complaint channels)","excerpt":null,"isParaphrase":false},{"shortCode":"OECD-AI-PRIN","name":"OECD AI Principles (Recommendation)","slug":"oecd-ai-principles","verdict":"governs","citation":"Principle 1.5 (accountability)","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"governs","citation":"Arts. 14-15 (procedural safeguards + remedies)","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF","name":"NIST AI Risk Management Framework","slug":"nist-ai-rmf","verdict":"implicit","citation":"Accountability characteristic","excerpt":"Post-deployment AI system monitoring plans are implemented, including mechanisms for capturing and evaluating input from users and other relevant AI actors, appeal and override, and decommissioning.","isParaphrase":true},{"shortCode":"NIST-AI-RMF-GENAI","name":"NIST AI RMF Generative AI Profile","slug":"nist-ai-rmf-genai-profile","verdict":"implicit","citation":"Accountability characteristic from base RMF; not GenAI-specific text","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-1047","name":"California SB-1047: Safe and Secure Innovation for Frontier AI Models Act","slug":"ca-sb-1047","verdict":"implicit","citation":"Whistleblower protections (§22607) + AG enforcement (§22608); no individual redress","excerpt":null,"isParaphrase":false},{"shortCode":"IN-DPDP-2023","name":"India Digital Personal Data Protection Act + AI Advisory (MEITY)","slug":"india-dpdpa","verdict":"governs","citation":"DPDPA §§13-15 (data principal rights, grievance + Data Protection Board)","excerpt":null,"isParaphrase":false},{"shortCode":"BR-AIBILL-2024","name":"Brazil AI Bill (PL 2338/2023)","slug":"brazil-ai-bill","verdict":"governs","citation":"PL 2338/2023 Art. 9 (right to contest AI decisions, ANPD as regulator)","excerpt":null,"isParaphrase":false},{"shortCode":"SG-MODEL-AI-2024","name":"Singapore Model AI Governance Framework for Generative AI","slug":"singapore-model-ai-governance","verdict":"implicit","citation":"Framework Dimension 1 (Accountability) + Dimension 4 (Incident Reporting); pairs with PDPA grievance regime","excerpt":null,"isParaphrase":false},{"shortCode":"JP-METI-AI-2024","name":"Japan METI AI Guidelines for Business","slug":"japan-meti-ai-guidelines","verdict":"implicit","citation":"Principle 6 (Accountability) + Principle 8 (Fair Competition) — sectoral redress channels assumed","excerpt":null,"isParaphrase":false},{"shortCode":"EU-GDPR-2016","name":"General Data Protection Regulation (GDPR)","slug":"gdpr","verdict":"governs","citation":"Art. 77 DPA complaint; Art. 79 effective judicial remedy; Art. 80 collective representation by NGOs; Art. 82 right to compensation; Art. 83 administrative fines","excerpt":null,"isParaphrase":false},{"shortCode":"OMB-M-24-10","name":"OMB Memorandum M-24-10 (Advancing Governance, Innovation, and Risk Management for Agency Use of AI)","slug":"omb-m-24-10","verdict":"governs","citation":"Attachment 1 §5(c)(v)(D) human consideration + remedy for rights-impacting AI; opt-out where practicable","excerpt":"For rights-impacting AI, agencies must provide timely human consideration and potential remedy through a fallback and escalation process where individuals can appeal or contest adverse decisions.","isParaphrase":true},{"shortCode":"GSA-AI-GUIDE-2024","name":"GSA Generative AI and Specialized Computing Infrastructure Acquisition Resource Guide","slug":"gsa-ai-acquisition-guide","verdict":"implicit","citation":"Guide references OMB M-24-10 Attachment 1 minimum practices including human-consideration + remedy for rights-impacting AI","excerpt":null,"isParaphrase":false},{"shortCode":"DOD-RAI-2022","name":"DoD Responsible AI Strategy and Implementation Pathway","slug":"dod-rai-strategy","verdict":"implicit","citation":"Ethical Principle 'Governable' — ability to disengage or deactivate; Tenet 2 calibrated reliance addresses operator-facing redress but not affected-civilian redress","excerpt":"…possessing the ability to detect and avoid unintended consequences, and the ability to disengage or deactivate deployed systems that demonstrate unintended behavior.","isParaphrase":false},{"shortCode":"FEDRAMP-AI-2024","name":"FedRAMP AI Cloud Procurement Guidance","slug":"fedramp-ai-guidance","verdict":"implicit","citation":"Guidance cross-walks to OMB M-24-10 minimum practices including human-consideration + remedy for rights-impacting AI","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-53","name":"California SB-53: Transparency in Frontier Artificial Intelligence Act (TFAIA)","slug":"ca-sb-53","verdict":"implicit","citation":"Lab. Code §§ 1107–1107.2 — whistleblower anti-retaliation gives covered employees a PRIVATE right of action (employee-brought civil suit, attorney's fees, injunctive relief); the substantive transparency/framework/incident obligations are AG-enforced only (§ 22757.15). No general consumer/data-subject redress for AI harms.","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-243","name":"California SB 243: Companion Chatbots","slug":"ca-sb-243","verdict":"governs","citation":"Cal. Bus. & Prof. Code § 22605 (added by SB 243) — private right of action: a person injured in fact by a violation may sue for injunctive relief, the greater of actual damages or $1,000 per violation, and attorney's fees and costs","excerpt":"A person who suffers injury in fact as a result of a violation of this chapter may bring a civil action to recover all of the following relief: (a) Injunctive relief. (b) Damages in an amount equal to the greater of actual damages or one thousand dollars ($1,000) per violation. (c) Reasonable attorney's fees and costs.","isParaphrase":false},{"shortCode":"EU-PLD-2024","name":"Revised Product Liability Directive (Directive (EU) 2024/2853)","slug":"eu-product-liability-directive","verdict":"governs","citation":"Arts. 6, 8, 9, 10 — strict-liability compensation for defective products incl. software/AI: compensable damage (Art. 6), liable economic operators (Art. 8), court-ordered evidence disclosure (Art. 9), and rebuttable presumptions of defect + causation (Art. 10)","excerpt":"A national court shall presume defectiveness or the causal link where the claimant faces excessive difficulties, in particular due to technical or scientific complexity, in proving it.","isParaphrase":true},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"governs","citation":"Policy Area 'Ethical governance and stewardship', para 55 — harms through AI investigated and redressed via enforcement + remedial actions","excerpt":"Member States should ensure that harms caused through AI systems are investigated and redressed, by enacting strong enforcement mechanisms and remedial actions","isParaphrase":false},{"shortCode":"EU-PWD-2024","name":"Directive (EU) 2024/2831 on improving working conditions in platform work","slug":"eu-platform-work-directive","verdict":"governs","citation":"Directive (EU) 2024/2831, Article 11","excerpt":"Article 11 gives platform workers a right to a written explanation of significant automated decisions and to human review and contestation, and provides that decisions to restrict, suspend or terminat","isParaphrase":true},{"shortCode":"CN-DEEPSYN-2022","name":"Provisions on the Administration of Deep Synthesis of Internet Information Services","slug":"china-deep-synthesis-provisions","verdict":"governs","citation":"Art. 12","excerpt":"设置便捷的用户申诉和公众投诉、举报入口，公布处理流程和反馈时限，及时受理、处理和反馈","isParaphrase":false},{"shortCode":"US-TAKEITDOWN-2025","name":"TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act)","slug":"take-it-down-act","verdict":"implicit","citation":"Pub. L. 119-12 — the 48-hour platform notice-and-removal process plus mandatory criminal restitution and forfeiture give nonconsensual-intimate-image / deepfake victims a targeted remedy; narrow to one harm domain and FTC-enforced with no private right of action","excerpt":null,"isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"implicit","citation":"No general right to contest AI decisions. Art. 4(3) gives a right to object to authorised processing of one's personal data; Art. 16(3)(b) delegates the Government to provide compensatory/injunctive remedies and sanctions for training-data violations; the deepfake offence (Art. 612-quater) is prosecuted on the victim's complaint.","excerpt":"… il diritto di opporsi ai trattamenti autorizzati dei propri dati personali. … [delega a] prevedere strumenti di tutela, di carattere risarcitorio o inibitorio …","isParaphrase":false},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"implicit","citation":"Act No. 53 of 2025, Art. 16","excerpt":"... analyze cases in which citizens' rights or interests have been infringed ... and ... provide guidance, advice and information to ... AI-utilizing business operators and other relevant persons ...","isParaphrase":true},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"implicit","citation":"GDC Objective 3, para 23(b) (A/RES/79/1, Annex I)","excerpt":"establishing effective oversight and remedy mechanisms.","isParaphrase":true}],"governsCount":13,"relatedConcepts":[{"code":"systemic-risk","label":"Systemic Risk (AI)","slug":"systemic-risk"},{"code":"model-card","label":"Model Card","slug":"model-card"},{"code":"scalable-oversight","label":"Scalable Oversight","slug":"scalable-oversight"},{"code":"training-data-attribution","label":"Training-Data Attribution","slug":"training-data-attribution"},{"code":"hallucination","label":"Hallucination","slug":"hallucination"},{"code":"retrieval-augmented-generation","label":"Retrieval-Augmented Generation (RAG)","slug":"retrieval-augmented-generation"}],"engagingArticles":["brazil-ai-bill","ca-sb-1047","ca-sb-243","ca-sb-53","china-deep-synthesis-provisions","china-genai-measures","coe-ai-convention","dod-rai-strategy","eu-ai-act","eu-platform-work-directive","eu-product-liability-directive","fedramp-ai-guidance","gdpr","gsa-ai-acquisition-guide","hallucination","india-dpdpa","italy-ai-law-2025","japan-ai-promotion-act","japan-meti-ai-guidelines","model-card","nist-ai-rmf","nist-ai-rmf-genai-profile","oecd-ai-principles","omb-m-24-10","redress","retrieval-augmented-generation","scalable-oversight","singapore-model-ai-governance","systemic-risk","take-it-down-act","training-data-attribution","uk-ai-white-paper","un-global-digital-compact","unesco-ai-ethics-recommendation"],"literatureCount":16,"convergence":[{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["brazil-ai-bill","ca-sb-1047","ca-sb-243","coe-ai-convention","dod-rai-strategy","eu-product-liability-directive","fedramp-ai-guidance","gdpr","gsa-ai-acquisition-guide","hallucination","india-dpdpa","japan-meti-ai-guidelines","model-card","nist-ai-rmf","oecd-ai-principles","omb-m-24-10","singapore-model-ai-governance","training-data-attribution","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":20},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["brazil-ai-bill","ca-sb-1047","ca-sb-53","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide","japan-meti-ai-guidelines","model-card","nist-ai-rmf","nist-ai-rmf-genai-profile","oecd-ai-principles","retrieval-augmented-generation","singapore-model-ai-governance","systemic-risk","uk-ai-white-paper"],"citeCount":16},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["ca-sb-1047","china-genai-measures","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","india-dpdpa","japan-meti-ai-guidelines","model-card","nist-ai-rmf","nist-ai-rmf-genai-profile","oecd-ai-principles","retrieval-augmented-generation","singapore-model-ai-governance","uk-ai-white-paper"],"citeCount":14},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["brazil-ai-bill","ca-sb-1047","ca-sb-53","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide","japan-meti-ai-guidelines","model-card","nist-ai-rmf-genai-profile","oecd-ai-principles","singapore-model-ai-governance","systemic-risk","uk-ai-white-paper"],"citeCount":14},{"key":"arxiv:2504.18236","title":null,"citingArticles":["ca-sb-1047","coe-ai-convention","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","hallucination","japan-meti-ai-guidelines","nist-ai-rmf","oecd-ai-principles","redress","singapore-model-ai-governance","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":13},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["ca-sb-1047","ca-sb-53","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide","omb-m-24-10","systemic-risk"],"citeCount":8},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["brazil-ai-bill","china-genai-measures","eu-ai-act","japan-meti-ai-guidelines","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance"],"citeCount":7},{"key":"lit-novelli-casolari-hacker-spedicato-floridi-generative-a","title":"Generative AI in EU law: Liability, privacy, intellectual property, and cybersecurity","citingArticles":["china-deep-synthesis-provisions","gdpr","hallucination","japan-meti-ai-guidelines","model-card","retrieval-augmented-generation","training-data-attribution"],"citeCount":7},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["gsa-ai-acquisition-guide","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","training-data-attribution"],"citeCount":5},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["china-genai-measures","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance"],"citeCount":5},{"key":"lit-gary-e-weissman-toni-mankowitz-genevieve-p-kanter-unre","title":"Unregulated large language models produce medical device-like output","citingArticles":["ca-sb-243","hallucination","omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["brazil-ai-bill","coe-ai-convention","india-dpdpa","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["ca-sb-1047","ca-sb-53","nist-ai-rmf","nist-ai-rmf-genai-profile","uk-ai-white-paper"],"citeCount":5},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["brazil-ai-bill","ca-sb-53","eu-product-liability-directive","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":5},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["brazil-ai-bill","china-genai-measures","india-dpdpa","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["china-deep-synthesis-provisions","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance"],"citeCount":5},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["brazil-ai-bill","ca-sb-53","eu-product-liability-directive","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":5},{"key":"lit-sarah-sterz-kevin-baum-sebastian-biewer-holger-hermann","title":"On the Quest for Effectiveness in Human Oversight: Interdisciplinary Perspectives","citingArticles":["ca-sb-243","gdpr","hallucination","model-card","retrieval-augmented-generation"],"citeCount":5},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["ca-sb-53","eu-product-liability-directive","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":4},{"key":"lit-eloundou-manning-mishkin-rock-gpts-are-gpts-labor-mark","title":"GPTs are GPTs: Labor market impact potential of LLMs","citingArticles":["fedramp-ai-guidance","gsa-ai-acquisition-guide","omb-m-24-10","retrieval-augmented-generation"],"citeCount":4},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["china-genai-measures","japan-meti-ai-guidelines","nist-ai-rmf","singapore-model-ai-governance"],"citeCount":4},{"key":"lit-sastry-heim-belfield-anderljung-brundage-et-al-computi","title":"Computing Power and the Governance of Artificial Intelligence","citingArticles":["ca-sb-1047","gsa-ai-acquisition-guide","omb-m-24-10","systemic-risk"],"citeCount":4},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["nist-ai-rmf-genai-profile","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["ca-sb-1047","ca-sb-53","dod-rai-strategy"],"citeCount":3},{"key":"lit-chiara-gallese-predictive-policing-and-predictive-just","title":"Predictive policing and predictive justice: Ethics, data protection, and the AI act","citingArticles":["coe-ai-convention","india-dpdpa","uk-ai-white-paper"],"citeCount":3},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["japan-meti-ai-guidelines","oecd-ai-principles","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["brazil-ai-bill","japan-meti-ai-guidelines","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-francesca-palmiotto-the-ai-act-roller-coaster-the-evol","title":"The AI Act Roller Coaster: The Evolution of Fundamental Rights Protection in the Legislative Process and the Future of the Regulation","citingArticles":["dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":3},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["fedramp-ai-guidance","gsa-ai-acquisition-guide","systemic-risk"],"citeCount":3},{"key":"lit-huijuan-peng-and-pey-woan-lee-reimagining-u-s-tort-law","title":"Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore","citingArticles":["nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance"],"citeCount":3},{"key":"lit-kyrie-zhixuan-zhou-abhinav-choudhry-ece-gumusel-and-ma","title":"'Sora is incredible and scary': public perceptions and governance challenges of text-to-video generative AI models","citingArticles":["china-genai-measures","japan-meti-ai-guidelines","singapore-model-ai-governance"],"citeCount":3},{"key":"lit-natalie-sheard-algorithm-facilitated-discrimination-a","title":"Algorithm-facilitated discrimination: a socio-legal study of the use by employers of artificial intelligence hiring systems","citingArticles":["omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["coe-ai-convention","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-wachter-mittelstadt-floridi-why-a-right-to-explanation","title":"Why a Right to Explanation of Automated Decision-Making Does Not Exist in the General Data Protection Regulation","citingArticles":["eu-platform-work-directive","gdpr","redress"],"citeCount":3},{"key":"lit-aditya-loganathan-michael-friedman-tayab-waseem-et-al","title":"Current state of Food and Drug Administration-approved artificial intelligence/machine learning medical devices: pathways, transparency, and evidence gaps","citingArticles":["hallucination","uk-ai-white-paper"],"citeCount":2},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["singapore-model-ai-governance","uk-ai-white-paper"],"citeCount":2},{"key":"lit-bengio-hinton-yao-song-et-al-managing-extreme-ai-risks","title":"Managing extreme AI risks amid rapid progress","citingArticles":["ca-sb-1047","dod-rai-strategy"],"citeCount":2},{"key":"lit-boris-babic-i-glenn-cohen-ariel-dora-stern-yiwen-li-me","title":"A general framework for governing marketed AI/ML medical devices","citingArticles":["ca-sb-243","hallucination"],"citeCount":2},{"key":"lit-emelie-stiernstr-mer-facial-recognition-technology-in","title":"Facial recognition technology in law enforcement: a scoping review of existing empirical studies","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-emre-bayaml-o-lu-the-right-to-contest-automated-decisi","title":"The right to contest automated decisions under the General Data Protection Regulation: Beyond the so-called 'right to explanation'","citingArticles":["eu-platform-work-directive","redress"],"citeCount":2},{"key":"lit-ezgi-yazici-toward-a-global-standard-for-ethical-ai-re","title":"Toward a global standard for ethical AI regulation: addressing gaps in AI-driven biometric and high-resolution satellite imaging in the EU AI Act","citingArticles":["fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":2},{"key":"lit-henrik-palmer-olsen-thomas-troels-hildebrandt-corneliu","title":"The Right to Transparency in Public Governance: Freedom of Information and the Use of Artificial Intelligence by Public Agencies","citingArticles":["ca-sb-243","model-card"],"citeCount":2},{"key":"lit-kars-alfrink-ianus-keller-gerd-kortuem-neelke-doorn-co","title":"Contestable AI by Design: Towards a Framework","citingArticles":["ca-sb-243","redress"],"citeCount":2},{"key":"lit-longpre-mahari-et-al-data-provenance-initiative-a-larg","title":"A large-scale audit of dataset licensing and attribution in AI","citingArticles":["retrieval-augmented-generation","training-data-attribution"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["japan-meti-ai-guidelines","singapore-model-ai-governance"],"citeCount":2},{"key":"lit-martin-kretschmer-bartolomeo-meletti-lionel-bently-gab","title":"Copyright and AI in the UK: Opting-In or Opting-Out?","citingArticles":["gsa-ai-acquisition-guide","retrieval-augmented-generation"],"citeCount":2},{"key":"lit-mary-phuong-matthew-aitchison-elliot-catt-et-al-google","title":"Evaluating Frontier Models for Dangerous Capabilities","citingArticles":["model-card","systemic-risk"],"citeCount":2},{"key":"lit-mireia-yurrita-tim-draws-agathe-balayn-dave-murray-rus","title":"Disentangling Fairness Perceptions in Algorithmic Decision-Making: the Effects of Explanations, Human Oversight, and Contestability","citingArticles":["ca-sb-243","redress"],"citeCount":2},{"key":"lit-petros-terzis-michael-veale-no-lle-gaumann-law-and-the","title":"Law and the Emerging Political Economy of Algorithmic Audits","citingArticles":["model-card","training-data-attribution"],"citeCount":2},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["ca-sb-1047","dod-rai-strategy"],"citeCount":2},{"key":"lit-sandra-fredman-darcy-du-toit-alessio-bertolini-jonas-v","title":"Fair Work for Platform Workers: Lessons from the EU Directive and Beyond","citingArticles":["eu-platform-work-directive","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-shai-farber-machines-of-justice-a-systematic-review-of","title":"Machines of justice: A systematic review of AI applications in policing and criminal justice","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-taner-kuru-lawfulness-of-the-mass-processing-of-public","title":"Lawfulness of the mass processing of publicly accessible online data to train large language models","citingArticles":["gdpr","retrieval-augmented-generation"],"citeCount":2},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["india-dpdpa","nist-ai-rmf-genai-profile"],"citeCount":2},{"key":"lit-vijaytha-muralidharan-madelena-y-ng-shada-alsalamah-sa","title":"Global Initiative on AI for Health (GI-AI4H): strategic priorities advancing governance across the United Nations","citingArticles":["ca-sb-243","hallucination"],"citeCount":2}]},{"topicCode":"international_coordination","label":"International Coordination","url":"https://policywindow.org/wiki/synthesis/international-coordination","themeArticle":"https://policywindow.org/wiki/international-coordination","empiricalConsensus":"emerging","contestedQuestion":"Will AI-governance coordination converge on the OECD / UN / G7 / GPAI / bilateral-MoU mode, or fragment into bloc-based regimes (US-led / EU-led / China-led)? Field consensus is forming but unsettled.","stances":[{"shortCode":"BLETCHLEY-2023","name":"Bletchley Declaration on AI Safety","slug":"bletchley-declaration","verdict":"governs","citation":"Declaration §8-10 (international coordination is the operative ask)","excerpt":null,"isParaphrase":false},{"shortCode":"SEOUL-2024","name":"Seoul Declaration on Safe, Innovative and Inclusive AI","slug":"seoul-declaration","verdict":"governs","citation":"Declaration §5-7 (AISI network, follow-up summits)","excerpt":null,"isParaphrase":false},{"shortCode":"ASEAN-AI-GUIDE-2024","name":"ASEAN Guide on AI Governance and Ethics","slug":"asean-ai-guide","verdict":"governs","citation":"Guide explicitly designed to harmonise across ASEAN-10 member states + interoperate with OECD AI Principles + G7 Hiroshima","excerpt":null,"isParaphrase":false},{"shortCode":"AU-AI-STRATEGY-2024","name":"African Union Continental AI Strategy","slug":"au-continental-ai-strategy","verdict":"governs","citation":"AU Strategy §6 (coordination with UN GA AI resolutions + AU-EU AI Working Group)","excerpt":null,"isParaphrase":false},{"shortCode":"ANTHROPIC-RSP-2024","name":"Anthropic Responsible Scaling Policy (RSP) v2","slug":"anthropic-rsp","verdict":"implicit","citation":"Seoul Frontier AI Safety Commitments signatory; coordinates with US + UK AISIs on capability evaluation","excerpt":null,"isParaphrase":false},{"shortCode":"OPENAI-PREPAREDNESS-2023","name":"OpenAI Preparedness Framework","slug":"openai-preparedness","verdict":"implicit","citation":"Seoul Frontier AI Safety Commitments signatory; pre-deployment evaluation sharing with US + UK AISIs","excerpt":null,"isParaphrase":false},{"shortCode":"DEEPMIND-FSF-2024","name":"Google DeepMind Frontier Safety Framework","slug":"deepmind-fsf","verdict":"implicit","citation":"Seoul Frontier AI Safety Commitments signatory; UK AISI pre-deployment evaluation cooperation","excerpt":null,"isParaphrase":false},{"shortCode":"META-FRONTIER-2024","name":"Meta Frontier AI Framework","slug":"meta-frontier-ai-framework","verdict":"implicit","citation":"Seoul Frontier AI Safety Commitments signatory","excerpt":null,"isParaphrase":false},{"shortCode":"UK-US-AISI-MOU-2024","name":"UK-US AI Safety Institute Memorandum of Understanding","slug":"uk-us-aisi-mou","verdict":"governs","citation":"MoU is the operative bilateral; precedent for the broader AISI network","excerpt":null,"isParaphrase":false},{"shortCode":"WH-VOLUNTARY-2023","name":"White House Voluntary AI Commitments","slug":"wh-voluntary-2023","verdict":"implicit","citation":"Precursor to Seoul Frontier AI Safety Commitments; same signatory base largely overlaps","excerpt":null,"isParaphrase":false},{"shortCode":"SG-MODEL-AI-2024","name":"Singapore Model AI Governance Framework for Generative AI","slug":"singapore-model-ai-governance","verdict":"governs","citation":"Framework explicitly aligns with G7 Hiroshima Code + OECD AI Principles; ASEAN Guide pairs","excerpt":null,"isParaphrase":false},{"shortCode":"JP-METI-AI-2024","name":"Japan METI AI Guidelines for Business","slug":"japan-meti-ai-guidelines","verdict":"governs","citation":"Guidelines explicit alignment with G7 Hiroshima AI Process Code of Conduct + OECD AI Principles","excerpt":null,"isParaphrase":false},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"governs","citation":"Policy Area 'Development and International Cooperation', para 80 — platforms for international cooperation on AI","excerpt":"Member States should work through international organizations to provide platforms for international cooperation on AI for development, including by contributing expertise, funding, data","isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"implicit","citation":"Art. 1(2)/Art. 2 align the law with EU Reg. 2024/1689; Art. 19(3) requires the national strategy to take account of international humanitarian law; Art. 20(2) designates ACN as the single contact point with EU institutions under AI-Act Art. 70.","excerpt":"La strategia … tiene conto dei princìpi del diritto internazionale umanitario … l'ACN è designata quale … punto di contatto unico con le istituzioni dell'Unione europea …","isParaphrase":false},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"governs","citation":"Act No. 53 of 2025, Arts. 17 & 3(5)","excerpt":"The State shall promote international cooperation in the research, development, and utilization of AI-related technology, and actively participate in the formulation of international norms in that field.","isParaphrase":true},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"governs","citation":"GDC Objective 5, paras 55(b) and 56 (A/RES/79/1, Annex I)","excerpt":"Support interoperability and compatibility of artificial intelligence governance approaches ...; Establish, within the United Nations, a multidisciplinary Independent International Scientific Panel on AI ...; Initiate ... a Global Dialogue on AI Governance.","isParaphrase":false}],"governsCount":10,"relatedConcepts":[{"code":"policy-instrument","label":"Policy Instrument","slug":"policy-instrument"},{"code":"hardware-enabled-governance","label":"Hardware-Enabled Governance Mechanisms","slug":"hardware-enabled-governance"}],"engagingArticles":["anthropic-rsp","asean-ai-guide","au-continental-ai-strategy","bletchley-declaration","deepmind-fsf","hardware-enabled-governance","international-coordination","italy-ai-law-2025","japan-ai-promotion-act","japan-meti-ai-guidelines","meta-frontier-ai-framework","openai-preparedness","policy-instrument","seoul-declaration","singapore-model-ai-governance","uk-us-aisi-mou","un-global-digital-compact","unesco-ai-ethics-recommendation","wh-voluntary-2023"],"literatureCount":10,"convergence":[{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["anthropic-rsp","asean-ai-guide","bletchley-declaration","deepmind-fsf","japan-meti-ai-guidelines","meta-frontier-ai-framework","openai-preparedness","policy-instrument","seoul-declaration","singapore-model-ai-governance","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":12},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["anthropic-rsp","bletchley-declaration","deepmind-fsf","japan-meti-ai-guidelines","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","singapore-model-ai-governance","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":10},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["anthropic-rsp","bletchley-declaration","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":8},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["anthropic-rsp","bletchley-declaration","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":8},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["asean-ai-guide","au-continental-ai-strategy","international-coordination","japan-meti-ai-guidelines","openai-preparedness","policy-instrument","singapore-model-ai-governance","uk-us-aisi-mou"],"citeCount":8},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["asean-ai-guide","bletchley-declaration","japan-meti-ai-guidelines","meta-frontier-ai-framework","singapore-model-ai-governance","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":7},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["anthropic-rsp","bletchley-declaration","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":7},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":6},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["asean-ai-guide","au-continental-ai-strategy","international-coordination","japan-meti-ai-guidelines","openai-preparedness","policy-instrument"],"citeCount":6},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["anthropic-rsp","bletchley-declaration","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","uk-us-aisi-mou"],"citeCount":6},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["anthropic-rsp","bletchley-declaration","openai-preparedness","policy-instrument","seoul-declaration","wh-voluntary-2023"],"citeCount":6},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["bletchley-declaration","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":6},{"key":"lit-maksym-andriushchenko-alexandra-souly-mateusz-dziemian","title":"AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents","citingArticles":["bletchley-declaration","deepmind-fsf","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":5},{"key":"lit-stephen-weymouth-digital-disintegration-techno-blocs-a","title":"Digital Disintegration: Techno-Blocs and Strategic Sovereignty in the AI Era","citingArticles":["asean-ai-guide","au-continental-ai-strategy","hardware-enabled-governance","international-coordination","policy-instrument"],"citeCount":5},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["anthropic-rsp","japan-meti-ai-guidelines","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":4},{"key":"arxiv:2504.18236","title":null,"citingArticles":["japan-meti-ai-guidelines","singapore-model-ai-governance","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["asean-ai-guide","au-continental-ai-strategy","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["asean-ai-guide","au-continental-ai-strategy","singapore-model-ai-governance"],"citeCount":3},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["au-continental-ai-strategy","japan-meti-ai-guidelines","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-david-m-allison-stephen-herzog-artificial-intelligence","title":"Artificial Intelligence and Nuclear Weapons Proliferation: The Technological Arms Race for (In)visibility","citingArticles":["bletchley-declaration","meta-frontier-ai-framework","openai-preparedness"],"citeCount":3},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["au-continental-ai-strategy","japan-meti-ai-guidelines","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["asean-ai-guide","au-continental-ai-strategy","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-kyrie-zhixuan-zhou-abhinav-choudhry-ece-gumusel-and-ma","title":"'Sora is incredible and scary': public perceptions and governance challenges of text-to-video generative AI models","citingArticles":["japan-meti-ai-guidelines","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":3},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["asean-ai-guide","au-continental-ai-strategy","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-megha-shrivastava-and-amrita-jash-china-s-semiconducto","title":"China's semiconductor conundrum: understanding US export controls and their efficacy","citingArticles":["anthropic-rsp","hardware-enabled-governance","meta-frontier-ai-framework"],"citeCount":3},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["japan-meti-ai-guidelines","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":3},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["japan-meti-ai-guidelines","singapore-model-ai-governance","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":2},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["hardware-enabled-governance","policy-instrument"],"citeCount":2},{"key":"lit-huijuan-peng-and-pey-woan-lee-reimagining-u-s-tort-law","title":"Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore","citingArticles":["singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":2},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":2},{"key":"lit-roberts-hine-taddeo-floridi-global-ai-governance-barri","title":"Global AI governance: barriers and pathways forward","citingArticles":["asean-ai-guide","international-coordination"],"citeCount":2},{"key":"lit-sastry-heim-belfield-anderljung-brundage-et-al-computi","title":"Computing Power and the Governance of Artificial Intelligence","citingArticles":["hardware-enabled-governance","policy-instrument"],"citeCount":2}]},{"topicCode":"national_security_carveouts","label":"National Security Carveouts in AI Regulation","url":"https://policywindow.org/wiki/synthesis/national-security-carveouts","themeArticle":"https://policywindow.org/wiki/national-security-carveouts","empiricalConsensus":"settled","contestedQuestion":"Whether the carveout should be (a) categorical exclusion of national-security AI, (b) parallel governance track with sui generis rules, or (c) full civilian-track compliance with national-security override. Most instruments choose (a); the field debates whether this leaves a dangerous gap.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Art. 2(3) explicitly excludes AI systems used exclusively for military, defence, or national-security purposes","excerpt":"This Regulation does not apply to AI systems where and in so far as they are placed on the market, put into service, or used with or without modification exclusively for military, defence or national security purposes…","isParaphrase":false},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"governs","citation":"§11 national-security exemption; NSM-10 parallel-track governance for national-security AI","excerpt":null,"isParaphrase":false},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"Defence + intelligence excluded via sectoral-regulator scope; carveout via omission rather than explicit clause","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"governs","citation":"Art. 3 — does not apply to AI used for national security / defence","excerpt":null,"isParaphrase":false},{"shortCode":"IN-DPDP-2023","name":"India Digital Personal Data Protection Act + AI Advisory (MEITY)","slug":"india-dpdpa","verdict":"implicit","citation":"DPDPA exemptions for state-security functions (Art. 17); not specifically AI but applies","excerpt":null,"isParaphrase":false},{"shortCode":"GSA-AI-GUIDE-2024","name":"GSA Generative AI and Specialized Computing Infrastructure Acquisition Resource Guide","slug":"gsa-ai-acquisition-guide","verdict":"implicit","citation":"Guide references existing federal supply-chain risk-management framework (FAR Part 4 Subpart 4.21) which carries national-security overlays","excerpt":null,"isParaphrase":false},{"shortCode":"DOD-RAI-2022","name":"DoD Responsible AI Strategy and Implementation Pathway","slug":"dod-rai-strategy","verdict":"governs","citation":"The S&IP IS the DoD-specific RAI framework; tenets + ethical principles operationalise the national-security AI use case rather than carving out from a civilian framework","excerpt":"DoD personnel will exercise appropriate levels of judgment and care, while remaining responsible for the development, deployment, and use of AI capabilities.","isParaphrase":false},{"shortCode":"FEDRAMP-AI-2024","name":"FedRAMP AI Cloud Procurement Guidance","slug":"fedramp-ai-guidance","verdict":"implicit","citation":"FedRAMP High baseline + JAB authorisation route exists for higher-sensitivity use cases; classified systems are outside FedRAMP scope and governed by separate ICD-503 / NIST SP 800-53 IC overlay frameworks","excerpt":null,"isParaphrase":false},{"shortCode":"DFARS-252-204","name":"DFARS Subpart 252.204 (Safeguarding Covered Defense Information and Cyber Incident Reporting)","slug":"dfars-252-204","verdict":"governs","citation":"252.204-7012 + CMMC clauses (-7019/-7020/-7021) are the operative national-security-overlay framework for defence-acquisition information security; the subpart IS the carveout regime","excerpt":"The Contractor shall provide adequate security on all covered contractor information systems … by implementing NIST Special Publication 800-171, Protecting Controlled Unclassified Information in Nonfederal Systems…","isParaphrase":true},{"shortCode":"CN-DEEPSYN-2022","name":"Provisions on the Administration of Deep Synthesis of Internet Information Services","slug":"china-deep-synthesis-provisions","verdict":"implicit","citation":"Art. 6, Art. 19 & Art. 20","excerpt":"Content prohibitions tied to national security/social order (Art. 6), filing (Art. 19), and security assessment (Art. 20) reflect a state-security orientation, but these are obligations imposed on pro","isParaphrase":true},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"governs","citation":"Art. 6 — activities for national-security purposes by the intelligence services, ACN cybersecurity/resilience, national-defence by the Armed Forces, and certain national-security policing are excluded from the law's scope (subject to fundamental-rights respect; further rules by regulation under l. 124/2007 art. 43).","excerpt":"[national-security, cybersecurity, national-defence and certain national-security policing activities] sono escluse dall'ambito applicativo della presente legge.","isParaphrase":true},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"implicit","citation":"Act No. 53 of 2025, Art. 3(2)","excerpt":"... They are also important technologies from the perspective of national security.","isParaphrase":true}],"governsCount":6,"relatedConcepts":[],"engagingArticles":["china-deep-synthesis-provisions","coe-ai-convention","dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide","india-dpdpa","italy-ai-law-2025","japan-ai-promotion-act","national-security-carveouts","uk-ai-white-paper","us-eo-14110"],"literatureCount":14,"convergence":[{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide","uk-ai-white-paper","us-eo-14110"],"citeCount":7},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["dfars-252-204","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","india-dpdpa","uk-ai-white-paper"],"citeCount":6},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide","uk-ai-white-paper"],"citeCount":6},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["dfars-252-204","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide","us-eo-14110"],"citeCount":6},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["coe-ai-convention","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","india-dpdpa","uk-ai-white-paper"],"citeCount":6},{"key":"arxiv:2504.18236","title":null,"citingArticles":["coe-ai-convention","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","uk-ai-white-paper"],"citeCount":5},{"key":"lit-francesca-palmiotto-the-ai-act-roller-coaster-the-evol","title":"The AI Act Roller Coaster: The Evolution of Fundamental Rights Protection in the Legislative Process and the Future of the Regulation","citingArticles":["dfars-252-204","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","national-security-carveouts"],"citeCount":5},{"key":"lit-chiara-gallese-predictive-policing-and-predictive-just","title":"Predictive policing and predictive justice: Ethics, data protection, and the AI act","citingArticles":["coe-ai-convention","india-dpdpa","national-security-carveouts","uk-ai-white-paper"],"citeCount":4},{"key":"lit-ezgi-yazici-toward-a-global-standard-for-ethical-ai-re","title":"Toward a global standard for ethical AI regulation: addressing gaps in AI-driven biometric and high-resolution satellite imaging in the EU AI Act","citingArticles":["dfars-252-204","fedramp-ai-guidance","gsa-ai-acquisition-guide","national-security-carveouts"],"citeCount":4},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["dfars-252-204","fedramp-ai-guidance","gsa-ai-acquisition-guide","us-eo-14110"],"citeCount":4},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["gsa-ai-acquisition-guide","india-dpdpa"],"citeCount":2},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["eu-ai-act","us-eo-14110"],"citeCount":2},{"key":"lit-eloundou-manning-mishkin-rock-gpts-are-gpts-labor-mark","title":"GPTs are GPTs: Labor market impact potential of LLMs","citingArticles":["fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":2},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["coe-ai-convention","india-dpdpa"],"citeCount":2},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["uk-ai-white-paper","us-eo-14110"],"citeCount":2},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["china-deep-synthesis-provisions","india-dpdpa"],"citeCount":2},{"key":"lit-novelli-casolari-hacker-spedicato-floridi-generative-a","title":"Generative AI in EU law: Liability, privacy, intellectual property, and cybersecurity","citingArticles":["china-deep-synthesis-provisions","dfars-252-204"],"citeCount":2},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["coe-ai-convention","uk-ai-white-paper"],"citeCount":2},{"key":"lit-sastry-heim-belfield-anderljung-brundage-et-al-computi","title":"Computing Power and the Governance of Artificial Intelligence","citingArticles":["dfars-252-204","gsa-ai-acquisition-guide"],"citeCount":2}]},{"topicCode":"open_weight_release","label":"Open-Weight Frontier Release","url":"https://policywindow.org/wiki/synthesis/open-weight-release","themeArticle":"https://policywindow.org/wiki/open-weight-release","empiricalConsensus":"contested","contestedQuestion":"Should frontier weight-release be governed by capability-tier (block above threshold) or by safety-evaluation-evidence (allow with pre-release red-team) or by recipient-restriction (export controls)? Three distinct frames currently in active conflict.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Art. 53(2) + Recital 102/104 — explicit open-source GPAI exemption (with caveats for systemic-risk models)","excerpt":"The obligations in paragraph 1, points (a) and (b), shall not apply to providers of AI models released under a free and open-source licence … unless they are general-purpose AI models with systemic risks.","isParaphrase":true},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"implicit","citation":"§4.6 NTIA report on dual-use foundation models specifically addresses open-weight risk; not binding obligation","excerpt":null,"isParaphrase":false},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"implicit","citation":"Art. 8 — registration / safety assessment applies regardless of weight release modality","excerpt":null,"isParaphrase":false},{"shortCode":"SEOUL-2024","name":"Seoul Declaration on Safe, Innovative and Inclusive AI","slug":"seoul-declaration","verdict":"implicit","citation":"Frontier AI Safety Commitments apply to all 16 signatories regardless of open/closed weight stance (Meta is signatory)","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-1047","name":"California SB-1047: Safe and Secure Innovation for Frontier AI Models Act","slug":"ca-sb-1047","verdict":"governs","citation":"Vetoed bill — would have required covered models (incl. open-weight releases) to adopt a safety & security protocol + self-certified compliance, with independent third-party audits from 2026 (Anthropic + Meta objected on different grounds)","excerpt":null,"isParaphrase":false},{"shortCode":"AU-AI-STRATEGY-2024","name":"African Union Continental AI Strategy","slug":"au-continental-ai-strategy","verdict":"implicit","citation":"Continental strategy frames AI capacity-building — open access to weights aligns with capacity goals","excerpt":null,"isParaphrase":false},{"shortCode":"ANTHROPIC-RSP-2024","name":"Anthropic Responsible Scaling Policy (RSP) v2","slug":"anthropic-rsp","verdict":"implicit","citation":"RSP applies to Anthropic's models which are closed-weight; framework does not address third-party open release","excerpt":null,"isParaphrase":false},{"shortCode":"OPENAI-PREPAREDNESS-2023","name":"OpenAI Preparedness Framework","slug":"openai-preparedness","verdict":"implicit","citation":"Framework applies to OpenAI deployments (closed-weight); does not address third-party open release","excerpt":null,"isParaphrase":false},{"shortCode":"DEEPMIND-FSF-2024","name":"Google DeepMind Frontier Safety Framework","slug":"deepmind-fsf","verdict":"implicit","citation":"Framework applies to Google DeepMind deployments (mostly closed); third-party open release not addressed","excerpt":null,"isParaphrase":false},{"shortCode":"META-FRONTIER-2024","name":"Meta Frontier AI Framework","slug":"meta-frontier-ai-framework","verdict":"governs","citation":"Framework's distinctive feature — explicit defence of open-weight release as governance posture; halt-training commitment if 'critical risk' threshold reached without mitigations","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-942","name":"California SB 942: AI Transparency Act","slug":"ca-sb-942","verdict":"governs","citation":"Cal. Bus. & Prof. Code § 22757.3(c) (added by SB 942, operative Aug. 2, 2026) — a covered provider that LICENSES its GenAI system to a third party must require by contract that the licensee preserve the § 22757.3(b) disclosure capability, and must revoke the license within 96 hours if the licensee disables it; reinforced by § 22757.3.2 (added by AB 853, operative Jan. 1, 2027), which bars a GenAI hosting platform distributing a system's source code or model weights from knowingly hosting a non-disclosing system","excerpt":"If a covered provider licenses its GenAI system to a third party, the covered provider shall require by contract that the licensee maintain the system's capability to include a disclosure required by subdivision (b) in content the system creates or alters.","isParaphrase":false},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"implicit","citation":"GDC Objective 5 capacity-building partnerships (A/RES/79/1, Annex I)","excerpt":"increase access to resources including open artificial intelligence models and systems, open training data and compute.","isParaphrase":false}],"governsCount":4,"relatedConcepts":[],"engagingArticles":["anthropic-rsp","au-continental-ai-strategy","ca-sb-1047","ca-sb-942","china-genai-measures","deepmind-fsf","eu-ai-act","meta-frontier-ai-framework","open-weight-release","openai-preparedness","seoul-declaration","un-global-digital-compact","us-eo-14110"],"literatureCount":11,"convergence":[{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["anthropic-rsp","ca-sb-1047","ca-sb-942","deepmind-fsf","eu-ai-act","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","us-eo-14110"],"citeCount":9},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["anthropic-rsp","ca-sb-1047","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","us-eo-14110"],"citeCount":7},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["anthropic-rsp","ca-sb-1047","deepmind-fsf","eu-ai-act","meta-frontier-ai-framework","openai-preparedness","seoul-declaration"],"citeCount":7},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["anthropic-rsp","ca-sb-1047","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration"],"citeCount":6},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["anthropic-rsp","ca-sb-1047","eu-ai-act","openai-preparedness","seoul-declaration","us-eo-14110"],"citeCount":6},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration"],"citeCount":5},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["anthropic-rsp","ca-sb-942","china-genai-measures","eu-ai-act","us-eo-14110"],"citeCount":5},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration"],"citeCount":5},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["ca-sb-1047","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness","seoul-declaration"],"citeCount":5},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["anthropic-rsp","deepmind-fsf","meta-frontier-ai-framework","openai-preparedness"],"citeCount":4},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["ca-sb-1047","china-genai-measures","meta-frontier-ai-framework"],"citeCount":3},{"key":"lit-maksym-andriushchenko-alexandra-souly-mateusz-dziemian","title":"AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents","citingArticles":["deepmind-fsf","openai-preparedness","seoul-declaration"],"citeCount":3},{"key":"lit-megha-shrivastava-and-amrita-jash-china-s-semiconducto","title":"China's semiconductor conundrum: understanding US export controls and their efficacy","citingArticles":["anthropic-rsp","meta-frontier-ai-framework","us-eo-14110"],"citeCount":3},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["ca-sb-942","china-genai-measures"],"citeCount":2},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["au-continental-ai-strategy","openai-preparedness"],"citeCount":2},{"key":"lit-david-m-allison-stephen-herzog-artificial-intelligence","title":"Artificial Intelligence and Nuclear Weapons Proliferation: The Technological Arms Race for (In)visibility","citingArticles":["meta-frontier-ai-framework","openai-preparedness"],"citeCount":2},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["au-continental-ai-strategy","china-genai-measures"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["au-continental-ai-strategy","openai-preparedness"],"citeCount":2},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["ca-sb-942","china-genai-measures"],"citeCount":2},{"key":"lit-sayash-kapoor-rishi-bommasani-kevin-klyman-shayne-long","title":"On the Societal Impact of Open Foundation Models","citingArticles":["ca-sb-942","open-weight-release"],"citeCount":2}]},{"topicCode":"sovereign_ai","label":"Sovereign AI Doctrine","url":"https://policywindow.org/wiki/synthesis/sovereign-ai","themeArticle":"https://policywindow.org/wiki/sovereign-ai","empiricalConsensus":"emerging","contestedQuestion":"Is jurisdiction-bound model deployment technically feasible at frontier scale? Field literature is sparse; doctrine is post-2023 and largely aspirational.","stances":[{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"governs","citation":"§4.2 (Commerce reporting on dual-use models + large compute clusters; IaaS rules)","excerpt":null,"isParaphrase":false},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"governs","citation":"Art. 17 (registration + algorithm filing)","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-53","name":"California SB-53: Transparency in Frontier Artificial Intelligence Act (TFAIA)","slug":"ca-sb-53","verdict":"implicit","citation":"Gov. Code § 11546.8 — CalCompute: a consortium to develop a framework for a public cloud computing cluster expanding access to compute (report due Jan. 1, 2027; operative on appropriation)","excerpt":null,"isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"implicit","citation":"No explicit sovereign-model/sovereign-compute mandate. Supported indirectly by Art. 5 (technological sovereignty + national-data-centre preference), Art. 19 (biennial national AI strategy, dual-use coordination with the Ministry of Defence) and Art. 23 (state investment in AI, cybersecurity and quantum computing).","excerpt":"… la sovranità tecnologica della Nazione … [national strategy] … [investimenti nei settori dell'intelligenza artificiale, della cybersicurezza e del calcolo quantistico] …","isParaphrase":true},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"implicit","citation":"Act No. 53 of 2025, Art. 3(2)","excerpt":"... maintaining Japan's capacity to conduct research and development of such technologies and enhancing the international competitiveness of related industries ...","isParaphrase":true}],"governsCount":2,"relatedConcepts":[{"code":"compute-threshold","label":"Compute Threshold (AI Governance)","slug":"compute-threshold"},{"code":"ai-supply-chain","label":"AI Supply Chain","slug":"ai-supply-chain"},{"code":"model-distillation-risk","label":"Model Distillation Risk","slug":"model-distillation-risk"}],"engagingArticles":["ai-supply-chain","ca-sb-53","china-genai-measures","compute-threshold","italy-ai-law-2025","japan-ai-promotion-act","model-distillation-risk","sovereign-ai","us-eo-14110"],"literatureCount":12,"convergence":[{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["ai-supply-chain","ca-sb-53","compute-threshold","model-distillation-risk","us-eo-14110"],"citeCount":5},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["ai-supply-chain","ca-sb-53","model-distillation-risk","us-eo-14110"],"citeCount":4},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["ai-supply-chain","compute-threshold","model-distillation-risk","us-eo-14110"],"citeCount":4},{"key":"lit-justin-kollar-andrew-stokols-geopolitical-ecologies-of","title":"Geopolitical ecologies of cloud capitalism: Territorial restructuring and the making of national computing power in the U.S. and China","citingArticles":["ai-supply-chain","ca-sb-53","compute-threshold","sovereign-ai"],"citeCount":4},{"key":"lit-sastry-heim-belfield-anderljung-brundage-et-al-computi","title":"Computing Power and the Governance of Artificial Intelligence","citingArticles":["ai-supply-chain","compute-threshold","model-distillation-risk","sovereign-ai"],"citeCount":4},{"key":"lit-stephen-weymouth-digital-disintegration-techno-blocs-a","title":"Digital Disintegration: Techno-Blocs and Strategic Sovereignty in the AI Era","citingArticles":["ai-supply-chain","compute-threshold","model-distillation-risk","sovereign-ai"],"citeCount":4},{"key":"lit-akash-r-wasil-tom-reed-jack-william-miller-peter-barne","title":"Verification methods for international AI agreements","citingArticles":["compute-threshold","model-distillation-risk"],"citeCount":2},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["china-genai-measures","us-eo-14110"],"citeCount":2},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["ai-supply-chain","china-genai-measures"],"citeCount":2},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["ca-sb-53","us-eo-14110"],"citeCount":2},{"key":"lit-lehdonvirta-w-hawkins-compute-north-vs-compute-south-t","title":"Compute North vs. Compute South: The Uneven Possibilities of Compute-based AI Governance Around the Globe","citingArticles":["ai-supply-chain","compute-threshold"],"citeCount":2},{"key":"lit-lennart-heim-tim-fist-janet-egan-sihao-huang-stephen-z","title":"Governing Through the Cloud: The Intermediary Role of Compute Providers in AI Regulation","citingArticles":["compute-threshold","model-distillation-risk"],"citeCount":2}]},{"topicCode":"synthetic_content_provenance","label":"Synthetic Content Provenance","url":"https://policywindow.org/wiki/synthesis/synthetic-content-provenance","themeArticle":"https://policywindow.org/wiki/synthetic-content-provenance","empiricalConsensus":"contested","contestedQuestion":"Should provenance be a model-provider obligation (watermark at generation), a platform obligation (label at distribution), or a recipient right (declare on request)? Each jurisdiction is currently selecting a different burden allocation.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Art. 50(2) — provider machine-readable marking obligation; Art. 50(4) — deployer disclosure for deep fakes (distinct from the `deepfakes` topic which focuses on misuse-harms)","excerpt":"Providers of AI systems generating synthetic audio, image, video or text shall ensure the outputs are marked in a machine-readable format and detectable as artificially generated or manipulated.","isParaphrase":true},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"governs","citation":"§4.5(a) — content authentication + watermarking standards via NIST + Commerce","excerpt":null,"isParaphrase":false},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"governs","citation":"Art. 12 — mandatory marking of generative-AI output; aligns with Deep Synthesis Rules (2022) tagging requirements","excerpt":null,"isParaphrase":false},{"shortCode":"G7-HIROSHIMA","name":"G7 Hiroshima AI Process Code of Conduct","slug":"g7-hiroshima-code","verdict":"governs","citation":"Code §6 — 'develop and deploy reliable content authentication and provenance mechanisms'","excerpt":null,"isParaphrase":false},{"shortCode":"UN-RES-2024","name":"UN GA Resolution on Safe, Secure, Trustworthy AI","slug":"un-ai-resolution-2024","verdict":"implicit","citation":"General call for state action on safe AI; provenance not specifically addressed","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF","name":"NIST AI Risk Management Framework","slug":"nist-ai-rmf","verdict":"implicit","citation":"General framework applies; provenance-specific guidance lives in the GenAI Profile","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF-GENAI","name":"NIST AI RMF Generative AI Profile","slug":"nist-ai-rmf-genai-profile","verdict":"governs","citation":"NIST AI 600-1 — Information Integrity is one of 12 named GenAI risk categories; covers synthetic-content labelling + provenance","excerpt":null,"isParaphrase":false},{"shortCode":"BR-AIBILL-2024","name":"Brazil AI Bill (PL 2338/2023)","slug":"brazil-ai-bill","verdict":"implicit","citation":"PL 2338 general accuracy + transparency obligations would extend to provenance via interpretation","excerpt":null,"isParaphrase":false},{"shortCode":"ANTHROPIC-RSP-2024","name":"Anthropic Responsible Scaling Policy (RSP) v2","slug":"anthropic-rsp","verdict":"implicit","citation":"Deployment-stage controls would include content provenance where capability tier requires","excerpt":null,"isParaphrase":false},{"shortCode":"WH-VOLUNTARY-2023","name":"White House Voluntary AI Commitments","slug":"wh-voluntary-2023","verdict":"governs","citation":"Voluntary commitment #5 — 'develop and deploy mechanisms that enable users to understand if audio or visual content is AI-generated, including robust provenance, watermarking, or both'","excerpt":null,"isParaphrase":false},{"shortCode":"SG-MODEL-AI-2024","name":"Singapore Model AI Governance Framework for Generative AI","slug":"singapore-model-ai-governance","verdict":"governs","citation":"Framework dimension 7 — Content Provenance (one of nine framework dimensions, paired with AI Verify Foundation's technical-testing toolkit)","excerpt":null,"isParaphrase":false},{"shortCode":"JP-METI-AI-2024","name":"Japan METI AI Guidelines for Business","slug":"japan-meti-ai-guidelines","verdict":"implicit","citation":"Principle 5 (Transparency) + Hiroshima-alignment imply provenance obligations via reference incorporation","excerpt":null,"isParaphrase":false},{"shortCode":"EU-GPAI-COP-2025","name":"EU General-Purpose AI Code of Practice","slug":"gpai-code-of-practice","verdict":"implicit","citation":"Chapter 1 transparency commitments brush against Art. 50(2) deployer marking + Art. 53(1)(a) provider documentation","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-942","name":"California SB 942: AI Transparency Act","slug":"ca-sb-942","verdict":"governs","citation":"Cal. Bus. & Prof. Code § 22757.3(b) (added by SB 942) — a covered provider must embed a machine-readable 'latent' disclosure in AI-generated image/video/audio conveying provenance metadata: provider name, GenAI system name and version, creation/alteration time, and a unique identifier; reinforced by § 22757.3.1 (AB 853, operative 2027) barring large online platforms from knowingly stripping system provenance data","excerpt":"A covered provider shall include a latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered provider's GenAI system","isParaphrase":false},{"shortCode":"CN-DEEPSYN-2022","name":"Provisions on the Administration of Deep Synthesis of Internet Information Services","slug":"china-deep-synthesis-provisions","verdict":"governs","citation":"Art. 16 & Art. 18","excerpt":"Art. 16: 采取技术措施添加……标识，并依照法律、行政法规和国家有关规定保存日志信息；Art. 18: 任何组织和个人不得采用技术手段删除、篡改、隐匿……深度合成标识","isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"implicit","citation":"No standalone watermarking/provenance-marking duty in the law itself; provenance is reached only indirectly — Art. 612-quater criminalises deceptive AI-altered media (turning on whether content is apt to deceive as to genuineness) and the general transparency principle (Art. 4). Content-marking duties are left to the EU AI Act (Art. 1(2)).","excerpt":"… immagini, video o voci falsificati o alterati mediante l'impiego di sistemi di intelligenza artificiale e idonei a indurre in inganno sulla loro genuinità …","isParaphrase":false},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"governs","citation":"GDC Objective 3, para 36(c) (A/RES/79/1, Annex I)","excerpt":"identification of artificial intelligence-generated material, authenticity certification for content and origins, labelling, watermarking and other techniques.","isParaphrase":false}],"governsCount":10,"relatedConcepts":[],"engagingArticles":["anthropic-rsp","brazil-ai-bill","ca-sb-942","china-deep-synthesis-provisions","china-genai-measures","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","italy-ai-law-2025","japan-meti-ai-guidelines","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","synthetic-content-provenance","un-ai-resolution-2024","un-global-digital-compact","us-eo-14110","wh-voluntary-2023"],"literatureCount":10,"convergence":[{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["anthropic-rsp","brazil-ai-bill","ca-sb-942","china-genai-measures","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","japan-meti-ai-guidelines","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","synthetic-content-provenance","un-ai-resolution-2024","us-eo-14110","wh-voluntary-2023"],"citeCount":15},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["anthropic-rsp","brazil-ai-bill","ca-sb-942","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","japan-meti-ai-guidelines","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","us-eo-14110","wh-voluntary-2023"],"citeCount":12},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["anthropic-rsp","brazil-ai-bill","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","japan-meti-ai-guidelines","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":9},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["ca-sb-942","china-genai-measures","g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":8},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["anthropic-rsp","g7-hiroshima-code","gpai-code-of-practice","nist-ai-rmf","nist-ai-rmf-genai-profile","un-ai-resolution-2024","us-eo-14110","wh-voluntary-2023"],"citeCount":8},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["ca-sb-942","china-genai-measures","japan-meti-ai-guidelines","nist-ai-rmf","singapore-model-ai-governance","synthetic-content-provenance","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":8},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["ca-sb-942","china-deep-synthesis-provisions","g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":7},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["china-genai-measures","japan-meti-ai-guidelines","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":6},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["anthropic-rsp","brazil-ai-bill","g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":5},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["anthropic-rsp","brazil-ai-bill","g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":5},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["ca-sb-942","g7-hiroshima-code","nist-ai-rmf-genai-profile","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":5},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["anthropic-rsp","g7-hiroshima-code","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":4},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["anthropic-rsp","gpai-code-of-practice","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":4},{"key":"lit-huijuan-peng-and-pey-woan-lee-reimagining-u-s-tort-law","title":"Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore","citingArticles":["nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":4},{"key":"lit-kyrie-zhixuan-zhou-abhinav-choudhry-ece-gumusel-and-ma","title":"'Sora is incredible and scary': public perceptions and governance challenges of text-to-video generative AI models","citingArticles":["china-genai-measures","japan-meti-ai-guidelines","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":4},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["anthropic-rsp","eu-ai-act","us-eo-14110","wh-voluntary-2023"],"citeCount":4},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["brazil-ai-bill","japan-meti-ai-guidelines","nist-ai-rmf","singapore-model-ai-governance"],"citeCount":4},{"key":"arxiv:2504.18236","title":null,"citingArticles":["japan-meti-ai-guidelines","nist-ai-rmf","singapore-model-ai-governance"],"citeCount":3},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["g7-hiroshima-code","nist-ai-rmf-genai-profile","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["g7-hiroshima-code","singapore-model-ai-governance","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["gpai-code-of-practice","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":3},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["brazil-ai-bill","japan-meti-ai-guidelines","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["brazil-ai-bill","china-genai-measures","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-alistair-knott-dino-pedreschi-raja-chatila-et-al-incl","title":"Generative AI models should include detection mechanisms as a condition for public release","citingArticles":["china-deep-synthesis-provisions","synthetic-content-provenance"],"citeCount":2},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["japan-meti-ai-guidelines","un-ai-resolution-2024"],"citeCount":2},{"key":"lit-groh-sankaranarayanan-singh-kim-lippman-picard-human-d","title":"Human detection of political speech deepfakes across transcripts, audio, and video","citingArticles":["ca-sb-942","china-deep-synthesis-provisions"],"citeCount":2},{"key":"lit-hanlin-zhang-benjamin-l-edelman-danilo-francati-daniel","title":"Watermarks in the Sand: Impossibility of Strong Watermarking for Generative Models","citingArticles":["china-deep-synthesis-provisions","synthetic-content-provenance"],"citeCount":2},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["brazil-ai-bill","un-ai-resolution-2024"],"citeCount":2},{"key":"lit-k-j-kevin-feng-nick-ritchie-pia-blumenthal-andy-parson","title":"Examining the Impact of Provenance-Enabled Media on Trust and Accuracy Perceptions","citingArticles":["china-deep-synthesis-provisions","synthetic-content-provenance"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["japan-meti-ai-guidelines","singapore-model-ai-governance"],"citeCount":2},{"key":"lit-megha-shrivastava-and-amrita-jash-china-s-semiconducto","title":"China's semiconductor conundrum: understanding US export controls and their efficacy","citingArticles":["anthropic-rsp","us-eo-14110"],"citeCount":2},{"key":"lit-novelli-casolari-hacker-spedicato-floridi-generative-a","title":"Generative AI in EU law: Liability, privacy, intellectual property, and cybersecurity","citingArticles":["china-deep-synthesis-provisions","japan-meti-ai-guidelines"],"citeCount":2}]},{"topicCode":"tech_sovereignty","label":"Technological Sovereignty","url":"https://policywindow.org/wiki/synthesis/tech-sovereignty","themeArticle":"https://policywindow.org/wiki/tech-sovereignty","empiricalConsensus":"emerging","contestedQuestion":"Can mid-sized economies sustain frontier-tier AI capability domestically, or does the compute-cost curve favour US/CN/EU only? Active debate in India, Brazil, ASEAN policy literatures.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"implicit","citation":"Recitals 1-5 + EU competence framing; AI Office establishes EU capacity","excerpt":"The purpose of this Regulation is to improve the functioning of the internal market and promote the uptake of human-centric and trustworthy artificial intelligence…","isParaphrase":false},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"governs","citation":"§5.3(b) + CHIPS Act overlap (BIS export controls, domestic compute)","excerpt":null,"isParaphrase":false},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"Sovereign-capability framing in UK AI Action Plan (2025) — not in 2023 white paper","excerpt":null,"isParaphrase":false},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"governs","citation":"Art. 4 + national-strategy alignment; domestic-AI doctrine explicit","excerpt":null,"isParaphrase":false},{"shortCode":"G7-HIROSHIMA","name":"G7 Hiroshima AI Process Code of Conduct","slug":"g7-hiroshima-code","verdict":"implicit","citation":"Adoption-by-developer framing; G7 carries implicit sovereignty assumptions","excerpt":null,"isParaphrase":false},{"shortCode":"UN-RES-2024","name":"UN GA Resolution on Safe, Secure, Trustworthy AI","slug":"un-ai-resolution-2024","verdict":"implicit","citation":"Calls for bridging digital divides — adjacent to but not sovereignty","excerpt":null,"isParaphrase":false},{"shortCode":"ASEAN-AI-GUIDE-2024","name":"ASEAN Guide on AI Governance and Ethics","slug":"asean-ai-guide","verdict":"implicit","citation":"Guide framing emphasises ASEAN-bloc capacity-building over external dependency","excerpt":null,"isParaphrase":false},{"shortCode":"AU-AI-STRATEGY-2024","name":"African Union Continental AI Strategy","slug":"au-continental-ai-strategy","verdict":"governs","citation":"AU Strategy §4 (continental compute + data infrastructure + skill-formation)","excerpt":null,"isParaphrase":false},{"shortCode":"SG-MODEL-AI-2024","name":"Singapore Model AI Governance Framework for Generative AI","slug":"singapore-model-ai-governance","verdict":"implicit","citation":"AI Verify Foundation positions Singapore as an interoperable AI-assurance hub","excerpt":null,"isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"governs","citation":"Art. 5 — the State must promote AI to raise national competitiveness and the 'technological sovereignty of the Nation' (¶1(a)) and may steer public e-procurement to favour solutions localising strategic data and disaster-recovery/business-continuity in national data centres (¶1(d)).","excerpt":"… al fine di accrescere la competitività del sistema economico nazionale e la sovranità tecnologica della Nazione nel quadro della strategia europea … privilegiate quelle soluzioni che garantiscono la localizzazione e l'elaborazione dei dati strategici presso data center posti nel territorio nazionale …","isParaphrase":false},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"implicit","citation":"Act No. 53 of 2025, Art. 3(2)","excerpt":"... maintaining Japan's capacity to conduct research and development of such technologies and enhancing the international competitiveness of related industries ... important technologies from the perspective of national security.","isParaphrase":true}],"governsCount":4,"relatedConcepts":[{"code":"inference-time-compute","label":"Inference-Time Compute","slug":"inference-time-compute"}],"engagingArticles":["asean-ai-guide","au-continental-ai-strategy","china-genai-measures","eu-ai-act","g7-hiroshima-code","inference-time-compute","italy-ai-law-2025","japan-ai-promotion-act","singapore-model-ai-governance","tech-sovereignty","uk-ai-white-paper","un-ai-resolution-2024","us-eo-14110"],"literatureCount":12,"convergence":[{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["asean-ai-guide","au-continental-ai-strategy","g7-hiroshima-code","singapore-model-ai-governance","tech-sovereignty","uk-ai-white-paper","un-ai-resolution-2024"],"citeCount":7},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["china-genai-measures","eu-ai-act","g7-hiroshima-code","singapore-model-ai-governance","un-ai-resolution-2024","us-eo-14110"],"citeCount":6},{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["asean-ai-guide","eu-ai-act","g7-hiroshima-code","singapore-model-ai-governance","uk-ai-white-paper","us-eo-14110"],"citeCount":6},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["asean-ai-guide","au-continental-ai-strategy","g7-hiroshima-code","un-ai-resolution-2024"],"citeCount":4},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["china-genai-measures","g7-hiroshima-code","singapore-model-ai-governance","un-ai-resolution-2024"],"citeCount":4},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["asean-ai-guide","china-genai-measures","singapore-model-ai-governance","uk-ai-white-paper"],"citeCount":4},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["g7-hiroshima-code","uk-ai-white-paper","un-ai-resolution-2024","us-eo-14110"],"citeCount":4},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["asean-ai-guide","au-continental-ai-strategy","china-genai-measures","un-ai-resolution-2024"],"citeCount":4},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["eu-ai-act","g7-hiroshima-code","singapore-model-ai-governance","uk-ai-white-paper"],"citeCount":4},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["asean-ai-guide","au-continental-ai-strategy","un-ai-resolution-2024"],"citeCount":3},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["asean-ai-guide","au-continental-ai-strategy","singapore-model-ai-governance"],"citeCount":3},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["eu-ai-act","inference-time-compute","us-eo-14110"],"citeCount":3},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["china-genai-measures","singapore-model-ai-governance","un-ai-resolution-2024"],"citeCount":3},{"key":"arxiv:2504.18236","title":null,"citingArticles":["singapore-model-ai-governance","uk-ai-white-paper"],"citeCount":2},{"key":"concept-inference-time-compute","title":"Inference-Time Compute","citingArticles":["inference-time-compute","tech-sovereignty"],"citeCount":2},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["asean-ai-guide","au-continental-ai-strategy"],"citeCount":2},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["au-continental-ai-strategy","un-ai-resolution-2024"],"citeCount":2},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["au-continental-ai-strategy","un-ai-resolution-2024"],"citeCount":2},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["inference-time-compute","us-eo-14110"],"citeCount":2},{"key":"lit-kyrie-zhixuan-zhou-abhinav-choudhry-ece-gumusel-and-ma","title":"'Sora is incredible and scary': public perceptions and governance challenges of text-to-video generative AI models","citingArticles":["china-genai-measures","singapore-model-ai-governance"],"citeCount":2},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["g7-hiroshima-code","singapore-model-ai-governance"],"citeCount":2},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["singapore-model-ai-governance","uk-ai-white-paper"],"citeCount":2},{"key":"lit-stephen-weymouth-digital-disintegration-techno-blocs-a","title":"Digital Disintegration: Techno-Blocs and Strategic Sovereignty in the AI Era","citingArticles":["asean-ai-guide","au-continental-ai-strategy"],"citeCount":2},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["g7-hiroshima-code","un-ai-resolution-2024"],"citeCount":2}]},{"topicCode":"training_data","label":"Training-Data Rights","url":"https://policywindow.org/wiki/synthesis/training-data","themeArticle":"https://policywindow.org/wiki/training-data","empiricalConsensus":"contested","contestedQuestion":"Does the EU CDSM Directive's TDM-exemption cover commercial foundation-model training? Major active litigation (NYT v OpenAI, Getty v Stability) and parallel claim regimes in UK/JP/US.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"implicit","citation":"Recital 105; CDSM Directive provides primary copyright framework","excerpt":"[providers of general-purpose AI models shall] draw up and make publicly available a sufficiently detailed summary about the content used for training of the general-purpose AI model…","isParaphrase":true},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"governs","citation":"Art. 7 (legal source + IP requirements)","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"implicit","citation":"Art. 11 (privacy + data protection)","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF","name":"NIST AI Risk Management Framework","slug":"nist-ai-rmf","verdict":"implicit","citation":"Manage 4: data integrity","excerpt":"Scientific integrity and TEVV considerations are identified and documented, including those related to experimental design, data collection and selection (e.g., availability, representativeness, suitability)…","isParaphrase":true},{"shortCode":"NIST-AI-RMF-GENAI","name":"NIST AI RMF Generative AI Profile","slug":"nist-ai-rmf-genai-profile","verdict":"governs","citation":"NIST AI 600-1 §3.4 Data Privacy + §3.7 Intellectual Property","excerpt":null,"isParaphrase":false},{"shortCode":"IN-DPDP-2023","name":"India Digital Personal Data Protection Act + AI Advisory (MEITY)","slug":"india-dpdpa","verdict":"governs","citation":"DPDPA §§4-7 (consent + purpose limitation for AI training data)","excerpt":null,"isParaphrase":false},{"shortCode":"BR-AIBILL-2024","name":"Brazil AI Bill (PL 2338/2023)","slug":"brazil-ai-bill","verdict":"implicit","citation":"PL 2338/2023 cross-references LGPD (2018) for data-rights baseline","excerpt":null,"isParaphrase":false},{"shortCode":"AU-AI-STRATEGY-2024","name":"African Union Continental AI Strategy","slug":"au-continental-ai-strategy","verdict":"implicit","citation":"AU Strategy §5 + Malabo Convention (2014) data-protection baseline","excerpt":null,"isParaphrase":false},{"shortCode":"META-FRONTIER-2024","name":"Meta Frontier AI Framework","slug":"meta-frontier-ai-framework","verdict":"implicit","citation":"Open-weight framing engages training-data + IP issues; not the framework's primary lane","excerpt":null,"isParaphrase":false},{"shortCode":"JP-METI-AI-2024","name":"Japan METI AI Guidelines for Business","slug":"japan-meti-ai-guidelines","verdict":"implicit","citation":"Principle 4 (Safety) + Principle 2 (Education-Literacy) brush against training-data norms; ACA copyright regime separately addresses","excerpt":null,"isParaphrase":false},{"shortCode":"EU-GDPR-2016","name":"General Data Protection Regulation (GDPR)","slug":"gdpr","verdict":"governs","citation":"Art. 5(1)(b) purpose limitation; Art. 6 lawful basis; Art. 9 special-category overlay for sensitive training data; Art. 5(1)(c) data minimisation","excerpt":null,"isParaphrase":false},{"shortCode":"EU-GPAI-COP-2025","name":"EU General-Purpose AI Code of Practice","slug":"gpai-code-of-practice","verdict":"governs","citation":"Chapter 2 (Copyright) — Art. 53(1)(c) training-data summary obligations + Art. 53(1)(d) text-and-data-mining opt-out compliance","excerpt":null,"isParaphrase":false},{"shortCode":"GSA-AI-GUIDE-2024","name":"GSA Generative AI and Specialized Computing Infrastructure Acquisition Resource Guide","slug":"gsa-ai-acquisition-guide","verdict":"implicit","citation":"Supply-chain risk-management considerations include training-data provenance + dependency disclosure","excerpt":null,"isParaphrase":false},{"shortCode":"FEDRAMP-AI-2024","name":"FedRAMP AI Cloud Procurement Guidance","slug":"fedramp-ai-guidance","verdict":"implicit","citation":"Supply-chain risk-management considerations include training-data + model-weight provenance disclosure within the SSP","excerpt":null,"isParaphrase":false},{"shortCode":"DFARS-252-204","name":"DFARS Subpart 252.204 (Safeguarding Covered Defense Information and Cyber Incident Reporting)","slug":"dfars-252-204","verdict":"governs","citation":"252.204-7012 — training-data sets stored on covered contractor information systems require NIST SP 800-171 implementation when designated CDI; data-spill / exfiltration events trigger 72-hour cyber-incident reporting under 252.204-7012(c)","excerpt":"When the Contractor discovers a cyber incident that affects a covered contractor information system … the Contractor shall … rapidly report cyber incidents to DoD … within 72 hours of discovery.","isParaphrase":true},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"governs","citation":"Policy Area 'Data Policy', para 71 — data-governance strategies ensuring continual evaluation of training-data quality","excerpt":"Member States should work to develop data governance strategies that ensure the continual evaluation of the quality of training data for AI systems","isParaphrase":false},{"shortCode":"CN-DEEPSYN-2022","name":"Provisions on the Administration of Deep Synthesis of Internet Information Services","slug":"china-deep-synthesis-provisions","verdict":"governs","citation":"Art. 14","excerpt":"深度合成服务提供者……应当加强训练数据管理，采取必要措施保障训练数据安全；训练数据包含个人信息的，应当遵守个人信息保护的有关规定","isParaphrase":false},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"governs","citation":"Art. 25 (new Art. 70-septies l. 633/1941) permits text-and-data-mining reproductions/extractions for AI training from lawfully accessible material (per Arts. 70-ter/70-quater); Art. 16 delegates the Government to enact an organic regime on data, algorithms and mathematical methods for training AI.","excerpt":"«Art. 70-septies … le riproduzioni e le estrazioni … ai fini dell'estrazione di testo e di dati attraverso modelli e sistemi di intelligenza artificiale, anche generativa, sono consentite in conformità alle disposizioni di cui agli articoli 70-ter e 70-quater».","isParaphrase":false},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"implicit","citation":"Act No. 53 of 2025, Arts. 12 & 3(4)","excerpt":"... intellectual infrastructure ... including datasets (meaning collections of information gathered for a specific purpose) ...","isParaphrase":true},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"implicit","citation":"GDC Objective 3 para 36(c) and Objective 5 capacity-building (A/RES/79/1, Annex I)","excerpt":"incorporation of safeguards into artificial intelligence model training processes ... open training data.","isParaphrase":false}],"governsCount":9,"relatedConcepts":[{"code":"dual-use-research-taxonomy","label":"Dual-Use Research Norms (DURC for AI)","slug":"dual-use-research-taxonomy"},{"code":"provenance-watermarking","label":"Provenance & Watermarking","slug":"provenance-watermarking"},{"code":"ai-supply-chain","label":"AI Supply Chain","slug":"ai-supply-chain"},{"code":"training-data-attribution","label":"Training-Data Attribution","slug":"training-data-attribution"},{"code":"data-poisoning","label":"Data Poisoning","slug":"data-poisoning"},{"code":"model-merging-risk","label":"Model-Merging Risk","slug":"model-merging-risk"},{"code":"retrieval-augmented-generation","label":"Retrieval-Augmented Generation (RAG)","slug":"retrieval-augmented-generation"}],"engagingArticles":["ai-supply-chain","au-continental-ai-strategy","brazil-ai-bill","china-deep-synthesis-provisions","china-genai-measures","coe-ai-convention","data-poisoning","dfars-252-204","dual-use-research-taxonomy","eu-ai-act","fedramp-ai-guidance","gdpr","gpai-code-of-practice","gsa-ai-acquisition-guide","india-dpdpa","italy-ai-law-2025","japan-ai-promotion-act","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-merging-risk","nist-ai-rmf","nist-ai-rmf-genai-profile","provenance-watermarking","retrieval-augmented-generation","training-data","training-data-attribution","un-global-digital-compact","unesco-ai-ethics-recommendation"],"literatureCount":27,"convergence":[{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["ai-supply-chain","brazil-ai-bill","data-poisoning","dfars-252-204","eu-ai-act","fedramp-ai-guidance","gpai-code-of-practice","gsa-ai-acquisition-guide","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-merging-risk","nist-ai-rmf","nist-ai-rmf-genai-profile","retrieval-augmented-generation"],"citeCount":14},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["ai-supply-chain","china-genai-measures","data-poisoning","dfars-252-204","fedramp-ai-guidance","gsa-ai-acquisition-guide","india-dpdpa","japan-meti-ai-guidelines","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","retrieval-augmented-generation"],"citeCount":12},{"key":"lit-novelli-casolari-hacker-spedicato-floridi-generative-a","title":"Generative AI in EU law: Liability, privacy, intellectual property, and cybersecurity","citingArticles":["ai-supply-chain","china-deep-synthesis-provisions","data-poisoning","dfars-252-204","dual-use-research-taxonomy","gdpr","japan-meti-ai-guidelines","model-merging-risk","provenance-watermarking","retrieval-augmented-generation","training-data","training-data-attribution"],"citeCount":12},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["ai-supply-chain","data-poisoning","gpai-code-of-practice","gsa-ai-acquisition-guide","india-dpdpa","meta-frontier-ai-framework","model-merging-risk","nist-ai-rmf","nist-ai-rmf-genai-profile","training-data","training-data-attribution"],"citeCount":11},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["brazil-ai-bill","data-poisoning","dfars-252-204","eu-ai-act","fedramp-ai-guidance","gpai-code-of-practice","gsa-ai-acquisition-guide","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-merging-risk","nist-ai-rmf-genai-profile"],"citeCount":11},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["brazil-ai-bill","coe-ai-convention","fedramp-ai-guidance","gdpr","gsa-ai-acquisition-guide","india-dpdpa","japan-meti-ai-guidelines","nist-ai-rmf","training-data-attribution","unesco-ai-ethics-recommendation"],"citeCount":10},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["brazil-ai-bill","china-genai-measures","eu-ai-act","gpai-code-of-practice","japan-meti-ai-guidelines","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":7},{"key":"arxiv:2504.18236","title":null,"citingArticles":["coe-ai-convention","fedramp-ai-guidance","gsa-ai-acquisition-guide","japan-meti-ai-guidelines","nist-ai-rmf","unesco-ai-ethics-recommendation"],"citeCount":6},{"key":"lit-martin-kretschmer-bartolomeo-meletti-lionel-bently-gab","title":"Copyright and AI in the UK: Opting-In or Opting-Out?","citingArticles":["ai-supply-chain","au-continental-ai-strategy","gpai-code-of-practice","gsa-ai-acquisition-guide","retrieval-augmented-generation","training-data"],"citeCount":6},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["au-continental-ai-strategy","brazil-ai-bill","coe-ai-convention","india-dpdpa","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-longpre-mahari-et-al-data-provenance-initiative-a-larg","title":"A large-scale audit of dataset licensing and attribution in AI","citingArticles":["data-poisoning","model-merging-risk","retrieval-augmented-generation","training-data","training-data-attribution"],"citeCount":5},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["au-continental-ai-strategy","brazil-ai-bill","china-genai-measures","india-dpdpa","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["china-deep-synthesis-provisions","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","provenance-watermarking"],"citeCount":5},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["ai-supply-chain","dfars-252-204","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":5},{"key":"lit-shayne-longpre-robert-mahari-ariel-lee-et-al-consent-i","title":"Consent in Crisis: The Rapid Decline of the AI Data Commons","citingArticles":["data-poisoning","dual-use-research-taxonomy","model-merging-risk","training-data","training-data-attribution"],"citeCount":5},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["china-genai-measures","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":4},{"key":"lit-eloundou-manning-mishkin-rock-gpts-are-gpts-labor-mark","title":"GPTs are GPTs: Labor market impact potential of LLMs","citingArticles":["dual-use-research-taxonomy","fedramp-ai-guidance","gsa-ai-acquisition-guide","retrieval-augmented-generation"],"citeCount":4},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["au-continental-ai-strategy","brazil-ai-bill","japan-meti-ai-guidelines","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["ai-supply-chain","dfars-252-204","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":4},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["gpai-code-of-practice","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":4},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["brazil-ai-bill","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":4},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["brazil-ai-bill","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":4},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":3},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["au-continental-ai-strategy","nist-ai-rmf-genai-profile","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["au-continental-ai-strategy","japan-meti-ai-guidelines","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-ezgi-yazici-toward-a-global-standard-for-ethical-ai-re","title":"Toward a global standard for ethical AI regulation: addressing gaps in AI-driven biometric and high-resolution satellite imaging in the EU AI Act","citingArticles":["dfars-252-204","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":3},{"key":"lit-francesca-palmiotto-the-ai-act-roller-coaster-the-evol","title":"The AI Act Roller Coaster: The Evolution of Fundamental Rights Protection in the Legislative Process and the Future of the Regulation","citingArticles":["dfars-252-204","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":3},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["china-genai-measures","japan-meti-ai-guidelines","nist-ai-rmf"],"citeCount":3},{"key":"lit-sarah-sterz-kevin-baum-sebastian-biewer-holger-hermann","title":"On the Quest for Effectiveness in Human Oversight: Interdisciplinary Perspectives","citingArticles":["data-poisoning","gdpr","retrieval-augmented-generation"],"citeCount":3},{"key":"lit-sastry-heim-belfield-anderljung-brundage-et-al-computi","title":"Computing Power and the Governance of Artificial Intelligence","citingArticles":["ai-supply-chain","dfars-252-204","gsa-ai-acquisition-guide"],"citeCount":3},{"key":"lit-taner-kuru-lawfulness-of-the-mass-processing-of-public","title":"Lawfulness of the mass processing of publicly accessible online data to train large language models","citingArticles":["gdpr","retrieval-augmented-generation","training-data"],"citeCount":3},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["india-dpdpa","nist-ai-rmf-genai-profile","provenance-watermarking"],"citeCount":3},{"key":"lit-anderljung-barnhart-korinek-et-al-frontier-ai-regulati","title":"Frontier AI Regulation: Managing Emerging Risks to Public Safety","citingArticles":["dual-use-research-taxonomy","model-merging-risk"],"citeCount":2},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["gpai-code-of-practice","meta-frontier-ai-framework"],"citeCount":2},{"key":"lit-beatriz-kira-when-non-consensual-intimate-deepfakes-go","title":"When non-consensual intimate deepfakes go viral: The insufficiency of the UK Online Safety Act","citingArticles":["china-deep-synthesis-provisions","provenance-watermarking"],"citeCount":2},{"key":"lit-chiara-gallese-predictive-policing-and-predictive-just","title":"Predictive policing and predictive justice: Ethics, data protection, and the AI act","citingArticles":["coe-ai-convention","india-dpdpa"],"citeCount":2},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["au-continental-ai-strategy","japan-meti-ai-guidelines"],"citeCount":2},{"key":"lit-groh-sankaranarayanan-singh-kim-lippman-picard-human-d","title":"Human detection of political speech deepfakes across transcripts, audio, and video","citingArticles":["china-deep-synthesis-provisions","provenance-watermarking"],"citeCount":2},{"key":"lit-huijuan-peng-and-pey-woan-lee-reimagining-u-s-tort-law","title":"Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore","citingArticles":["nist-ai-rmf","nist-ai-rmf-genai-profile"],"citeCount":2},{"key":"lit-kaigeng-li-hong-wu-yupeng-dong-copyright-protection-du","title":"Copyright protection during the training stage of generative AI: Industry-oriented U.S. law, rights-oriented EU law, and fair remuneration rights for generative AI training under the UN's international governance regime for AI","citingArticles":["training-data","training-data-attribution"],"citeCount":2},{"key":"lit-keith-raymond-harris-ai-or-your-lying-eyes-some-shortc","title":"AI or Your Lying Eyes: Some Shortcomings of Artificially Intelligent Deepfake Detectors","citingArticles":["china-deep-synthesis-provisions","provenance-watermarking"],"citeCount":2},{"key":"lit-kyrie-zhixuan-zhou-abhinav-choudhry-ece-gumusel-and-ma","title":"'Sora is incredible and scary': public perceptions and governance challenges of text-to-video generative AI models","citingArticles":["china-genai-measures","japan-meti-ai-guidelines"],"citeCount":2},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["au-continental-ai-strategy","japan-meti-ai-guidelines"],"citeCount":2},{"key":"lit-mary-phuong-matthew-aitchison-elliot-catt-et-al-google","title":"Evaluating Frontier Models for Dangerous Capabilities","citingArticles":["dual-use-research-taxonomy","model-merging-risk"],"citeCount":2},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["coe-ai-convention","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-petros-terzis-michael-veale-no-lle-gaumann-law-and-the","title":"Law and the Emerging Political Economy of Algorithmic Audits","citingArticles":["data-poisoning","training-data-attribution"],"citeCount":2},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["gpai-code-of-practice","meta-frontier-ai-framework"],"citeCount":2},{"key":"lit-stephen-weymouth-digital-disintegration-techno-blocs-a","title":"Digital Disintegration: Techno-Blocs and Strategic Sovereignty in the AI Era","citingArticles":["ai-supply-chain","au-continental-ai-strategy"],"citeCount":2}]},{"topicCode":"transparency","label":"Transparency Obligations","url":"https://policywindow.org/wiki/synthesis/transparency","themeArticle":"https://policywindow.org/wiki/transparency","empiricalConsensus":"contested","contestedQuestion":"Does transparency disclosure (model cards, training-data summaries) actually reduce bias / misuse / accidents? Selbst & Barocas (2019) argue disclosure ≠ fairness; regulators assume it helps.","stances":[{"shortCode":"EU-AIA-2024","name":"EU AI Act","slug":"eu-ai-act","verdict":"governs","citation":"Arts. 13, 50 (transparency obligations)","excerpt":"Providers shall ensure that AI systems intended to interact directly with natural persons … are informed that they are interacting with an AI system.","isParaphrase":false},{"shortCode":"US-EO-14110","name":"Executive Order 14110 on Safe, Secure, Trustworthy AI","slug":"us-eo-14110","verdict":"implicit","citation":"§4.2(a)(i) (reporting includes red-team results)","excerpt":null,"isParaphrase":false},{"shortCode":"UK-WHITEPAPER-2023","name":"UK Pro-Innovation Approach to AI Regulation (White Paper)","slug":"uk-ai-white-paper","verdict":"implicit","citation":"Principle 4 (transparency + explainability)","excerpt":null,"isParaphrase":false},{"shortCode":"CN-GENAI-2023","name":"Interim Measures for Generative AI Service Management","slug":"china-genai-measures","verdict":"conflicts","citation":"Art. 4 + Algorithm Recommendation Rules — disclosure to CAC, not public; conflicts with EU public-disclosure model","excerpt":null,"isParaphrase":false},{"shortCode":"G7-HIROSHIMA","name":"G7 Hiroshima AI Process Code of Conduct","slug":"g7-hiroshima-code","verdict":"governs","citation":"Code §2 (publicly report capabilities, limitations)","excerpt":null,"isParaphrase":false},{"shortCode":"OECD-AI-PRIN","name":"OECD AI Principles (Recommendation)","slug":"oecd-ai-principles","verdict":"governs","citation":"Principle 1.3 (transparency + explainability)","excerpt":null,"isParaphrase":false},{"shortCode":"COE-AI-CONV","name":"Council of Europe Framework Convention on AI","slug":"coe-ai-convention","verdict":"governs","citation":"Art. 8 (transparency + oversight)","excerpt":null,"isParaphrase":false},{"shortCode":"UN-RES-2024","name":"UN GA Resolution on Safe, Secure, Trustworthy AI","slug":"un-ai-resolution-2024","verdict":"implicit","citation":"Calls for trustworthy AI broadly","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF","name":"NIST AI Risk Management Framework","slug":"nist-ai-rmf","verdict":"governs","citation":"Trustworthy characteristics 5 (transparency) + 6 (explainability)","excerpt":"The AI model is explained, validated, and documented, and AI system output is interpreted within its context — as identified in the MAP function — to inform responsible use and governance.","isParaphrase":true},{"shortCode":"BLETCHLEY-2023","name":"Bletchley Declaration on AI Safety","slug":"bletchley-declaration","verdict":"implicit","citation":"Declaration §6 endorses transparency to evaluators; no operative requirements","excerpt":null,"isParaphrase":false},{"shortCode":"SEOUL-2024","name":"Seoul Declaration on Safe, Innovative and Inclusive AI","slug":"seoul-declaration","verdict":"governs","citation":"Declaration §4 + Commitments §3 (publish safety frameworks)","excerpt":null,"isParaphrase":false},{"shortCode":"NIST-AI-RMF-GENAI","name":"NIST AI RMF Generative AI Profile","slug":"nist-ai-rmf-genai-profile","verdict":"governs","citation":"Govern + Map cross-cutting documentation requirements applied to GenAI","excerpt":null,"isParaphrase":false},{"shortCode":"CA-SB-1047","name":"California SB-1047: Safe and Secure Innovation for Frontier AI Models Act","slug":"ca-sb-1047","verdict":"implicit","citation":"Required safety determinations are public; full safety case is to regulator only","excerpt":null,"isParaphrase":false},{"shortCode":"IN-DPDP-2023","name":"India Digital Personal Data Protection Act + AI Advisory (MEITY)","slug":"india-dpdpa","verdict":"implicit","citation":"DPDPA §5 notice requirements + MEITY Mar-2024 Advisory transparency mandates","excerpt":null,"isParaphrase":false},{"shortCode":"BR-AIBILL-2024","name":"Brazil AI Bill (PL 2338/2023)","slug":"brazil-ai-bill","verdict":"governs","citation":"PL 2338/2023 Art. 7 (right to information about AI use + algorithmic explanation)","excerpt":null,"isParaphrase":false},{"shortCode":"ASEAN-AI-GUIDE-2024","name":"ASEAN Guide on AI Governance and Ethics","slug":"asean-ai-guide","verdict":"governs","citation":"ASEAN Guide §4 (transparency + explainability principle)","excerpt":null,"isParaphrase":false},{"shortCode":"ANTHROPIC-RSP-2024","name":"Anthropic Responsible Scaling Policy (RSP) v2","slug":"anthropic-rsp","verdict":"governs","citation":"RSP v2 §5 — public publication of safety determinations + capability eval methodology","excerpt":null,"isParaphrase":false},{"shortCode":"OPENAI-PREPAREDNESS-2023","name":"OpenAI Preparedness Framework","slug":"openai-preparedness","verdict":"implicit","citation":"Public Preparedness Reports + Safety Advisory Group decisions; full evaluation methodology partially disclosed","excerpt":null,"isParaphrase":false},{"shortCode":"DEEPMIND-FSF-2024","name":"Google DeepMind Frontier Safety Framework","slug":"deepmind-fsf","verdict":"implicit","citation":"FSF publication discloses framework + thresholds; per-evaluation outputs not consistently public","excerpt":null,"isParaphrase":false},{"shortCode":"META-FRONTIER-2024","name":"Meta Frontier AI Framework","slug":"meta-frontier-ai-framework","verdict":"governs","citation":"Open-weight release + framework publication is itself a transparency posture; trade-off discussed in framework text","excerpt":null,"isParaphrase":false},{"shortCode":"UK-US-AISI-MOU-2024","name":"UK-US AI Safety Institute Memorandum of Understanding","slug":"uk-us-aisi-mou","verdict":"implicit","citation":"Information sharing between AISIs; not public-facing transparency obligations","excerpt":null,"isParaphrase":false},{"shortCode":"WH-VOLUNTARY-2023","name":"White House Voluntary AI Commitments","slug":"wh-voluntary-2023","verdict":"governs","citation":"Commitments §6 (public reporting on capabilities, limitations, appropriate use)","excerpt":null,"isParaphrase":false},{"shortCode":"SG-MODEL-AI-2024","name":"Singapore Model AI Governance Framework for Generative AI","slug":"singapore-model-ai-governance","verdict":"governs","citation":"Framework Dimension 7 (Content Provenance) + Dimension 5 (Testing + Assurance) — pairs with AI Verify toolkit","excerpt":null,"isParaphrase":false},{"shortCode":"JP-METI-AI-2024","name":"Japan METI AI Guidelines for Business","slug":"japan-meti-ai-guidelines","verdict":"governs","citation":"Guidelines Principle 5 (Transparency) — model documentation + capability disclosure","excerpt":null,"isParaphrase":false},{"shortCode":"EU-GDPR-2016","name":"General Data Protection Regulation (GDPR)","slug":"gdpr","verdict":"governs","citation":"Arts. 12-14 (information to data subjects); Art. 13(2)(f) + 14(2)(g) meaningful information about ADM logic; Art. 22(3) suitable safeguards","excerpt":null,"isParaphrase":false},{"shortCode":"EU-GPAI-COP-2025","name":"EU General-Purpose AI Code of Practice","slug":"gpai-code-of-practice","verdict":"governs","citation":"Chapter 1 (Transparency) — 13 commitments + ~40 measures operationalising Art. 53(1)(a)-(c) model documentation + training-data summary","excerpt":null,"isParaphrase":false},{"shortCode":"OMB-M-24-10","name":"OMB Memorandum M-24-10 (Advancing Governance, Innovation, and Risk Management for Agency Use of AI)","slug":"omb-m-24-10","verdict":"governs","citation":"§3(a)(iv) public AI use-case inventory; Attachment 1 §5(c)(v) plain-language public notice + explanation for rights-impacting AI","excerpt":"Agencies must individually inventory each of their AI use cases at least annually, submit the inventory to OMB, and post a public version of the inventory on the agency website.","isParaphrase":true},{"shortCode":"GSA-AI-GUIDE-2024","name":"GSA Generative AI and Specialized Computing Infrastructure Acquisition Resource Guide","slug":"gsa-ai-acquisition-guide","verdict":"governs","citation":"Due-diligence questions call for vendor disclosure of training-data provenance, evaluation results, and model documentation","excerpt":"Faithful summary: the guide's due-diligence questions direct agencies to seek vendor disclosure of training-data provenance, evaluation and benchmarking results, and model documentation as part of AI acquisition.","isParaphrase":true},{"shortCode":"DOD-RAI-2022","name":"DoD Responsible AI Strategy and Implementation Pathway","slug":"dod-rai-strategy","verdict":"governs","citation":"Ethical Principle 'Traceable' + Tenet 2 (Warfighter Trust) — documentation + explainability requirements integrated into T&E + V&V lifecycle","excerpt":"The Department's AI capabilities will be developed and deployed such that relevant personnel possess an appropriate understanding of the technology, development processes, and operational methods…","isParaphrase":false},{"shortCode":"FEDRAMP-AI-2024","name":"FedRAMP AI Cloud Procurement Guidance","slug":"fedramp-ai-guidance","verdict":"governs","citation":"FedRAMP authorisation requires System Security Plan + control documentation; GenAI guidance extends to vendor disclosure of training-data provenance, evaluation results, model documentation","excerpt":"Faithful summary: FedRAMP authorisation requires a System Security Plan documenting NIST SP 800-53 controls; GenAI guidance extends disclosure to training-data provenance, evaluation results, and model documentation.","isParaphrase":true},{"shortCode":"CA-SB-53","name":"California SB-53: Transparency in Frontier Artificial Intelligence Act (TFAIA)","slug":"ca-sb-53","verdict":"governs","citation":"Bus. & Prof. Code § 22757.12 — frontier developers must publish a frontier AI framework + a pre-deployment transparency report","excerpt":"Before, or concurrently with, deploying a new or substantially modified frontier model, a frontier developer shall clearly and conspicuously publish on its internet website a transparency report…","isParaphrase":true},{"shortCode":"CA-SB-243","name":"California SB 243: Companion Chatbots","slug":"ca-sb-243","verdict":"governs","citation":"Cal. Bus. & Prof. Code § 22602(a) (added by SB 243) — operator must issue a clear-and-conspicuous notification that the companion chatbot is artificially generated and not human where a reasonable person would be misled; § 22602(c) adds, for known minors, a default every-three-hours AI-reminder + break notification","excerpt":"If a reasonable person interacting with a companion chatbot would be misled to believe that the person is interacting with a human, an operator shall issue a clear and conspicuous notification indicating that the companion chatbot is artificially generated and not human.","isParaphrase":false},{"shortCode":"CA-SB-942","name":"California SB 942: AI Transparency Act","slug":"ca-sb-942","verdict":"governs","citation":"Cal. Bus. & Prof. Code § 22757.2(a) (added by SB 942) — a covered provider must make available, free and publicly accessible, an AI detection tool that lets a user assess whether image/video/audio content was created or altered by that provider's GenAI system; reinforced by § 22757.3(a) manifest-disclosure user option","excerpt":"A covered provider shall make available an AI detection tool at no cost to the user that meets all of the following criteria","isParaphrase":false},{"shortCode":"EU-PLD-2024","name":"Revised Product Liability Directive (Directive (EU) 2024/2853)","slug":"eu-product-liability-directive","verdict":"implicit","citation":"Art. 9 — court-ordered disclosure of relevant evidence in the defendant's control, reinforced by the Art. 10(2)(a) adverse presumption for non-disclosure","excerpt":null,"isParaphrase":false},{"shortCode":"UNESCO-AI-ETHICS-2021","name":"UNESCO Recommendation on the Ethics of Artificial Intelligence","slug":"unesco-ai-ethics-recommendation","verdict":"governs","citation":"Principle 'Transparency and explainability', para 38 — people informed of AI-based decisions + right to request explanation","excerpt":"People should be fully informed when a decision is informed by or is made on the basis of AI algorithms... and should have the opportunity to request explanatory information","isParaphrase":false},{"shortCode":"EU-PWD-2024","name":"Directive (EU) 2024/2831 on improving working conditions in platform work","slug":"eu-platform-work-directive","verdict":"governs","citation":"Directive (EU) 2024/2831, Article 9 (with Arts. 7-8)","excerpt":"Article 9 requires digital labour platforms to inform persons performing platform work and their representatives about the use, categories, parameters and effects of automated monitoring systems and a","isParaphrase":true},{"shortCode":"CN-DEEPSYN-2022","name":"Provisions on the Administration of Deep Synthesis of Internet Information Services","slug":"china-deep-synthesis-provisions","verdict":"governs","citation":"Art. 16 & Art. 17","excerpt":"Art. 16: 对使用其服务生成或者编辑的信息内容，应当采取技术措施添加不影响用户使用的标识；Art. 17: 应当……进行显著标识，向公众提示深度合成情况","isParaphrase":false},{"shortCode":"NY-RAISE-2025","name":"New York RAISE Act: Responsible AI Safety and Education Act","slug":"ny-raise-act","verdict":"governs","citation":"N.Y. Gen. Bus. Law § 1421(1)(C) — a large developer must conspicuously publish (with appropriate redactions) its written safety and security protocol and transmit a copy to the attorney general","excerpt":"[A large developer shall] conspicuously publish a copy of its safety and security protocol with appropriate redactions and transmit a copy of such redacted protocol to the attorney general.","isParaphrase":true},{"shortCode":"IT-AILAW-2025","name":"Italy Law No. 132/2025 on Artificial Intelligence (Legge 23 settembre 2025, n. 132)","slug":"italy-ai-law-2025","verdict":"governs","citation":"Multiple operative disclosure duties: Art. 4(3) clear-language information on AI data processing + right to object; Art. 7(3) patient information; Art. 11(2) worker notification; Art. 13(2) professional's duty to disclose AI use to the client.","excerpt":"Le informazioni e le comunicazioni relative al trattamento dei dati … sono rese con linguaggio chiaro e semplice, in modo da garantire all'utente la conoscibilità dei relativi rischi e il diritto di opporsi …","isParaphrase":false},{"shortCode":"JP-AIPROMO-2025","name":"Japan AI Promotion Act (Act on the Promotion of Research, Development and Utilization of AI-Related Technologies)","slug":"japan-ai-promotion-act","verdict":"governs","citation":"Act No. 53 of 2025, Art. 3(4)","excerpt":"... necessary measures to ensure proper implementation, including securing transparency in the processes of such research, development, and utilization ...","isParaphrase":true},{"shortCode":"UN-GDC-2024","name":"UN Global Digital Compact","slug":"un-global-digital-compact","verdict":"governs","citation":"GDC Objective 5, para 55(d) (A/RES/79/1, Annex I)","excerpt":"Promote transparency, accountability and robust human oversight of artificial intelligence systems in compliance with international law (all SDGs).","isParaphrase":false}],"governsCount":30,"relatedConcepts":[{"code":"designated-systemic","label":"Designated Systemic-Risk Model","slug":"designated-systemic"},{"code":"model-card","label":"Model Card","slug":"model-card"},{"code":"alignment","label":"AI Alignment","slug":"alignment"},{"code":"scalable-oversight","label":"Scalable Oversight","slug":"scalable-oversight"},{"code":"capability-elicitation","label":"Capability Elicitation","slug":"capability-elicitation"},{"code":"dual-use-research-taxonomy","label":"Dual-Use Research Norms (DURC for AI)","slug":"dual-use-research-taxonomy"},{"code":"provenance-watermarking","label":"Provenance & Watermarking","slug":"provenance-watermarking"},{"code":"policy-instrument","label":"Policy Instrument","slug":"policy-instrument"},{"code":"training-data-attribution","label":"Training-Data Attribution","slug":"training-data-attribution"},{"code":"prompt-injection","label":"Prompt Injection","slug":"prompt-injection"},{"code":"agentic-system","label":"Agentic AI System","slug":"agentic-system"},{"code":"multi-turn-evaluation","label":"Multi-Turn Evaluation","slug":"multi-turn-evaluation"},{"code":"data-poisoning","label":"Data Poisoning","slug":"data-poisoning"},{"code":"jailbreak-resistance","label":"Jailbreak Resistance","slug":"jailbreak-resistance"},{"code":"hallucination","label":"Hallucination","slug":"hallucination"},{"code":"in-context-learning","label":"In-Context Learning","slug":"in-context-learning"},{"code":"retrieval-augmented-generation","label":"Retrieval-Augmented Generation (RAG)","slug":"retrieval-augmented-generation"}],"engagingArticles":["agentic-system","alignment","anthropic-rsp","asean-ai-guide","bletchley-declaration","brazil-ai-bill","ca-sb-1047","ca-sb-243","ca-sb-53","ca-sb-942","capability-elicitation","china-deep-synthesis-provisions","china-genai-measures","coe-ai-convention","data-poisoning","deepmind-fsf","designated-systemic","dod-rai-strategy","dual-use-research-taxonomy","eu-ai-act","eu-platform-work-directive","eu-product-liability-directive","fedramp-ai-guidance","g7-hiroshima-code","gdpr","gpai-code-of-practice","gsa-ai-acquisition-guide","hallucination","in-context-learning","india-dpdpa","italy-ai-law-2025","jailbreak-resistance","japan-ai-promotion-act","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-card","multi-turn-evaluation","nist-ai-rmf","nist-ai-rmf-genai-profile","ny-raise-act","oecd-ai-principles","omb-m-24-10","openai-preparedness","policy-instrument","prompt-injection","provenance-watermarking","retrieval-augmented-generation","scalable-oversight","seoul-declaration","singapore-model-ai-governance","training-data-attribution","transparency","uk-ai-white-paper","uk-us-aisi-mou","un-ai-resolution-2024","un-global-digital-compact","unesco-ai-ethics-recommendation","us-eo-14110","wh-voluntary-2023"],"literatureCount":27,"convergence":[{"key":"lit-david-fern-ndez-llorca-emilia-g-mez-ignacio-s-nchez-ga","title":"An interdisciplinary account of the terminological choices by EU policymakers ahead of the final agreement on the AI Act: AI system, general purpose AI system, foundation model, and generative AI","citingArticles":["anthropic-rsp","asean-ai-guide","bletchley-declaration","brazil-ai-bill","ca-sb-1047","ca-sb-53","ca-sb-942","capability-elicitation","data-poisoning","deepmind-fsf","designated-systemic","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","g7-hiroshima-code","gpai-code-of-practice","gsa-ai-acquisition-guide","in-context-learning","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-card","multi-turn-evaluation","nist-ai-rmf","nist-ai-rmf-genai-profile","oecd-ai-principles","openai-preparedness","policy-instrument","retrieval-augmented-generation","seoul-declaration","singapore-model-ai-governance","uk-ai-white-paper","uk-us-aisi-mou","us-eo-14110","wh-voluntary-2023"],"citeCount":34},{"key":"lit-martina-hulok-the-eu-model-of-ai-governance-regulating","title":"The EU model of AI governance: regulating artificial intelligence through law and policy","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-1047","ca-sb-53","capability-elicitation","data-poisoning","deepmind-fsf","designated-systemic","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","g7-hiroshima-code","gpai-code-of-practice","gsa-ai-acquisition-guide","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-card","multi-turn-evaluation","nist-ai-rmf-genai-profile","oecd-ai-principles","openai-preparedness","seoul-declaration","singapore-model-ai-governance","uk-ai-white-paper","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":27},{"key":"lit-hannah-ruschemeier-generative-ai-and-data-protection","title":"Generative AI and data protection","citingArticles":["asean-ai-guide","bletchley-declaration","ca-sb-1047","china-genai-measures","data-poisoning","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","india-dpdpa","japan-meti-ai-guidelines","meta-frontier-ai-framework","model-card","nist-ai-rmf","nist-ai-rmf-genai-profile","oecd-ai-principles","retrieval-augmented-generation","singapore-model-ai-governance","uk-ai-white-paper","uk-us-aisi-mou","wh-voluntary-2023"],"citeCount":20},{"key":"lit-mireia-yurrita-himanshu-verma-agathe-balayn-kars-alfri","title":"Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability","citingArticles":["brazil-ai-bill","ca-sb-1047","ca-sb-243","coe-ai-convention","dod-rai-strategy","eu-product-liability-directive","fedramp-ai-guidance","gdpr","gsa-ai-acquisition-guide","hallucination","india-dpdpa","japan-meti-ai-guidelines","model-card","nist-ai-rmf","oecd-ai-principles","omb-m-24-10","singapore-model-ai-governance","training-data-attribution","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":20},{"key":"lit-kirolos-eskandar-artificial-intelligence-and-synthetic","title":"Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","deepmind-fsf","g7-hiroshima-code","gpai-code-of-practice","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-ai-white-paper","uk-us-aisi-mou","un-ai-resolution-2024","us-eo-14110","wh-voluntary-2023"],"citeCount":17},{"key":"lit-matteo-pistillo-pablo-villalobos-defending-compute-thr","title":"Defending Compute Thresholds Against Legal Loopholes","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","capability-elicitation","designated-systemic","dod-rai-strategy","eu-ai-act","fedramp-ai-guidance","gsa-ai-acquisition-guide","in-context-learning","omb-m-24-10","openai-preparedness","policy-instrument","seoul-declaration","us-eo-14110","wh-voluntary-2023"],"citeCount":17},{"key":"lit-bram-rijsbosch-gijs-van-dijck-and-konrad-kollnig-missi","title":"Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act","citingArticles":["anthropic-rsp","brazil-ai-bill","ca-sb-942","china-genai-measures","eu-ai-act","g7-hiroshima-code","gpai-code-of-practice","japan-meti-ai-guidelines","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","un-ai-resolution-2024","us-eo-14110","wh-voluntary-2023"],"citeCount":14},{"key":"lit-atoosa-kasirzadeh-two-types-of-ai-existential-risk-dec","title":"Two types of AI existential risk: decisive and accumulative","citingArticles":["anthropic-rsp","bletchley-declaration","ca-sb-1047","ca-sb-53","deepmind-fsf","dod-rai-strategy","gpai-code-of-practice","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":13},{"key":"lit-noam-kolt-governing-ai-agents","title":"Governing AI Agents","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","eu-product-liability-directive","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":13},{"key":"arxiv:2504.18236","title":null,"citingArticles":["ca-sb-1047","coe-ai-convention","dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide","hallucination","japan-meti-ai-guidelines","nist-ai-rmf","oecd-ai-principles","singapore-model-ai-governance","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":12},{"key":"lit-lewis-hammond-alan-chan-jesse-clifton-et-al-cooperativ","title":"Multi-Agent Risks from Advanced AI","citingArticles":["anthropic-rsp","bletchley-declaration","brazil-ai-bill","ca-sb-53","deepmind-fsf","eu-product-liability-directive","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","uk-us-aisi-mou"],"citeCount":12},{"key":"lit-alan-chan-kevin-wei-sihao-huang-nitarshan-rajkumar-eli","title":"Infrastructure for AI Agents","citingArticles":["anthropic-rsp","ca-sb-53","deepmind-fsf","eu-product-liability-directive","g7-hiroshima-code","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":11},{"key":"lit-novelli-casolari-hacker-spedicato-floridi-generative-a","title":"Generative AI in EU law: Liability, privacy, intellectual property, and cybersecurity","citingArticles":["china-deep-synthesis-provisions","data-poisoning","dual-use-research-taxonomy","gdpr","hallucination","japan-meti-ai-guidelines","model-card","provenance-watermarking","retrieval-augmented-generation","training-data-attribution"],"citeCount":10},{"key":"lit-bao-kham-chau-and-george-he-audio-deepfakes-and-the-re","title":"Audio deepfakes and the regulation of the landlords of creativity","citingArticles":["ca-sb-942","china-genai-measures","g7-hiroshima-code","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":9},{"key":"lit-eloundou-manning-mishkin-rock-gpts-are-gpts-labor-mark","title":"GPTs are GPTs: Labor market impact potential of LLMs","citingArticles":["ca-sb-942","capability-elicitation","dual-use-research-taxonomy","fedramp-ai-guidance","gsa-ai-acquisition-guide","in-context-learning","omb-m-24-10","policy-instrument","retrieval-augmented-generation"],"citeCount":9},{"key":"lit-mateusz-abuz-a-teleological-interpretation-of-the-defi","title":"A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing'","citingArticles":["ca-sb-942","china-deep-synthesis-provisions","g7-hiroshima-code","india-dpdpa","nist-ai-rmf","nist-ai-rmf-genai-profile","provenance-watermarking","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":9},{"key":"lit-rebecca-scholefield-samuel-martin-otto-barten-internat","title":"International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty","citingArticles":["bletchley-declaration","ca-sb-1047","deepmind-fsf","dod-rai-strategy","gpai-code-of-practice","meta-frontier-ai-framework","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":9},{"key":"lit-arne-radeisen-open-foundation-models-and-tdm-exception","title":"Open Foundation Models and TDM Exceptions to Copyright – Building Blocks for an AI Ecosystem","citingArticles":["data-poisoning","gpai-code-of-practice","gsa-ai-acquisition-guide","india-dpdpa","meta-frontier-ai-framework","nist-ai-rmf","nist-ai-rmf-genai-profile","training-data-attribution"],"citeCount":8},{"key":"lit-heim-koessler-training-compute-thresholds-features-and","title":"Training Compute Thresholds: Features and Functions in AI Regulation","citingArticles":["capability-elicitation","designated-systemic","fedramp-ai-guidance","gsa-ai-acquisition-guide","in-context-learning","policy-instrument","us-eo-14110"],"citeCount":7},{"key":"lit-huw-roberts-mariarosaria-taddeo-luciano-floridi-a-fram","title":"A Framework for Evaluating Global AI Governance Initiatives","citingArticles":["asean-ai-guide","brazil-ai-bill","coe-ai-convention","india-dpdpa","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":7},{"key":"lit-maarten-buyl-alexander-rogiers-sander-noels-et-al-larg","title":"Large language models reflect the ideology of their creators","citingArticles":["asean-ai-guide","brazil-ai-bill","china-genai-measures","india-dpdpa","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":7},{"key":"lit-mary-phuong-matthew-aitchison-elliot-catt-et-al-google","title":"Evaluating Frontier Models for Dangerous Capabilities","citingArticles":["capability-elicitation","designated-systemic","dual-use-research-taxonomy","in-context-learning","jailbreak-resistance","model-card","multi-turn-evaluation"],"citeCount":7},{"key":"lit-mimi-zou-and-lu-zhang-navigating-china-s-regulatory-ap","title":"Navigating China's regulatory approach to generative artificial intelligence and large language models","citingArticles":["ca-sb-942","china-genai-measures","japan-meti-ai-guidelines","nist-ai-rmf","singapore-model-ai-governance","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":7},{"key":"lit-sarah-sterz-kevin-baum-sebastian-biewer-holger-hermann","title":"On the Quest for Effectiveness in Human Oversight: Interdisciplinary Perspectives","citingArticles":["ca-sb-243","data-poisoning","gdpr","hallucination","model-card","multi-turn-evaluation","retrieval-augmented-generation"],"citeCount":7},{"key":"lit-valentine-ugwuoke-and-madelyn-rose-sanfilippo-the-curr","title":"The Current Landscape of Deepfake Legislation in the United States","citingArticles":["ca-sb-942","g7-hiroshima-code","india-dpdpa","nist-ai-rmf-genai-profile","provenance-watermarking","un-ai-resolution-2024","wh-voluntary-2023"],"citeCount":7},{"key":"lit-anderljung-barnhart-korinek-et-al-frontier-ai-regulati","title":"Frontier AI Regulation: Managing Emerging Risks to Public Safety","citingArticles":["capability-elicitation","dual-use-research-taxonomy","model-card","multi-turn-evaluation","policy-instrument","us-eo-14110"],"citeCount":6},{"key":"lit-andr-ebert-joseph-alder-ralf-herbrich-philipp-hacker-a","title":"AI, Climate, and Regulation: From Data Centers to the AI Act","citingArticles":["asean-ai-guide","g7-hiroshima-code","nist-ai-rmf-genai-profile","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":6},{"key":"lit-mark-robinson-the-establishment-of-an-international-ai","title":"The establishment of an international AI agency: an applied solution to global AI governance","citingArticles":["asean-ai-guide","japan-meti-ai-guidelines","openai-preparedness","policy-instrument","singapore-model-ai-governance","uk-us-aisi-mou"],"citeCount":6},{"key":"lit-sastry-heim-belfield-anderljung-brundage-et-al-computi","title":"Computing Power and the Governance of Artificial Intelligence","citingArticles":["ca-sb-1047","capability-elicitation","designated-systemic","gsa-ai-acquisition-guide","omb-m-24-10","policy-instrument"],"citeCount":6},{"key":"lit-andreas-baur-european-ambitions-captured-by-american-c","title":"European ambitions captured by American clouds: digital sovereignty through Gaia-X?","citingArticles":["asean-ai-guide","g7-hiroshima-code","singapore-model-ai-governance","uk-ai-white-paper","un-ai-resolution-2024"],"citeCount":5},{"key":"lit-gary-e-weissman-toni-mankowitz-genevieve-p-kanter-unre","title":"Unregulated large language models produce medical device-like output","citingArticles":["ca-sb-243","hallucination","omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":5},{"key":"lit-henrik-palmer-olsen-thomas-troels-hildebrandt-corneliu","title":"The Right to Transparency in Public Governance: Freedom of Information and the Use of Artificial Intelligence by Public Agencies","citingArticles":["ca-sb-243","in-context-learning","model-card","multi-turn-evaluation","transparency"],"citeCount":5},{"key":"lit-maksym-andriushchenko-alexandra-souly-mateusz-dziemian","title":"AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents","citingArticles":["bletchley-declaration","deepmind-fsf","openai-preparedness","seoul-declaration","uk-us-aisi-mou"],"citeCount":5},{"key":"lit-petros-terzis-michael-veale-no-lle-gaumann-law-and-the","title":"Law and the Emerging Political Economy of Algorithmic Audits","citingArticles":["data-poisoning","model-card","multi-turn-evaluation","training-data-attribution","transparency"],"citeCount":5},{"key":"lit-council-of-europe-introductory-note-by-marc-rotenberg","title":"Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council Eur.) — with Introductory Note","citingArticles":["asean-ai-guide","japan-meti-ai-guidelines","openai-preparedness","policy-instrument"],"citeCount":4},{"key":"lit-daron-acemoglu-the-simple-macroeconomics-of-ai","title":"The simple macroeconomics of AI","citingArticles":["japan-meti-ai-guidelines","oecd-ai-principles","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"lit-david-m-allison-stephen-herzog-artificial-intelligence","title":"Artificial Intelligence and Nuclear Weapons Proliferation: The Technological Arms Race for (In)visibility","citingArticles":["bletchley-declaration","dod-rai-strategy","meta-frontier-ai-framework","openai-preparedness"],"citeCount":4},{"key":"lit-erik-brynjolfsson-danielle-li-and-lindsey-r-raymond-ge","title":"Generative AI at Work","citingArticles":["brazil-ai-bill","japan-meti-ai-guidelines","un-ai-resolution-2024","unesco-ai-ethics-recommendation"],"citeCount":4},{"key":"lit-huijuan-peng-and-pey-woan-lee-reimagining-u-s-tort-law","title":"Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore","citingArticles":["nist-ai-rmf","nist-ai-rmf-genai-profile","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":4},{"key":"lit-kyrie-zhixuan-zhou-abhinav-choudhry-ece-gumusel-and-ma","title":"'Sora is incredible and scary': public perceptions and governance challenges of text-to-video generative AI models","citingArticles":["china-genai-measures","japan-meti-ai-guidelines","singapore-model-ai-governance","wh-voluntary-2023"],"citeCount":4},{"key":"lit-anka-reuel-ben-bucknall-stephen-casper-tim-fist-lennar","title":"Open Problems in Technical AI Governance","citingArticles":["capability-elicitation","in-context-learning","multi-turn-evaluation"],"citeCount":3},{"key":"lit-bengio-hinton-yao-song-et-al-managing-extreme-ai-risks","title":"Managing extreme AI risks amid rapid progress","citingArticles":["ca-sb-1047","dod-rai-strategy","uk-us-aisi-mou"],"citeCount":3},{"key":"lit-chiara-gallese-predictive-policing-and-predictive-just","title":"Predictive policing and predictive justice: Ethics, data protection, and the AI act","citingArticles":["coe-ai-convention","india-dpdpa","uk-ai-white-paper"],"citeCount":3},{"key":"lit-francesca-palmiotto-the-ai-act-roller-coaster-the-evol","title":"The AI Act Roller Coaster: The Evolution of Fundamental Rights Protection in the Legislative Process and the Future of the Regulation","citingArticles":["dod-rai-strategy","fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":3},{"key":"lit-groh-sankaranarayanan-singh-kim-lippman-picard-human-d","title":"Human detection of political speech deepfakes across transcripts, audio, and video","citingArticles":["ca-sb-942","china-deep-synthesis-provisions","provenance-watermarking"],"citeCount":3},{"key":"lit-longpre-mahari-et-al-data-provenance-initiative-a-larg","title":"A large-scale audit of dataset licensing and attribution in AI","citingArticles":["data-poisoning","retrieval-augmented-generation","training-data-attribution"],"citeCount":3},{"key":"lit-martin-kretschmer-bartolomeo-meletti-lionel-bently-gab","title":"Copyright and AI in the UK: Opting-In or Opting-Out?","citingArticles":["gpai-code-of-practice","gsa-ai-acquisition-guide","retrieval-augmented-generation"],"citeCount":3},{"key":"lit-megha-shrivastava-and-amrita-jash-china-s-semiconducto","title":"China's semiconductor conundrum: understanding US export controls and their efficacy","citingArticles":["anthropic-rsp","meta-frontier-ai-framework","us-eo-14110"],"citeCount":3},{"key":"lit-natalie-sheard-algorithm-facilitated-discrimination-a","title":"Algorithm-facilitated discrimination: a socio-legal study of the use by employers of artificial intelligence hiring systems","citingArticles":["omb-m-24-10","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-pedro-robles-daniel-j-mallinson-eric-best-cheryl-devan","title":"Global perspectives on regulating facial recognition technology utilization for criminal justice arrests","citingArticles":["coe-ai-convention","uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":3},{"key":"lit-shayne-longpre-robert-mahari-ariel-lee-et-al-consent-i","title":"Consent in Crisis: The Rapid Decline of the AI Data Commons","citingArticles":["data-poisoning","dual-use-research-taxonomy","training-data-attribution"],"citeCount":3},{"key":"lit-wachter-mittelstadt-floridi-why-a-right-to-explanation","title":"Why a Right to Explanation of Automated Decision-Making Does Not Exist in the General Data Protection Regulation","citingArticles":["eu-platform-work-directive","gdpr","transparency"],"citeCount":3},{"key":"arxiv:1606.06565","title":null,"citingArticles":["alignment","scalable-oversight"],"citeCount":2},{"key":"arxiv:1706.03741","title":null,"citingArticles":["alignment","scalable-oversight"],"citeCount":2},{"key":"concept-prompt-injection","title":"Prompt Injection","citingArticles":["prompt-injection","retrieval-augmented-generation"],"citeCount":2},{"key":"lit-aditya-loganathan-michael-friedman-tayab-waseem-et-al","title":"Current state of Food and Drug Administration-approved artificial intelligence/machine learning medical devices: pathways, transparency, and evidence gaps","citingArticles":["hallucination","uk-ai-white-paper"],"citeCount":2},{"key":"lit-akash-r-wasil-tom-reed-jack-william-miller-peter-barne","title":"Verification methods for international AI agreements","citingArticles":["capability-elicitation","designated-systemic"],"citeCount":2},{"key":"lit-beatriz-kira-when-non-consensual-intimate-deepfakes-go","title":"When non-consensual intimate deepfakes go viral: The insufficiency of the UK Online Safety Act","citingArticles":["china-deep-synthesis-provisions","provenance-watermarking"],"citeCount":2},{"key":"lit-boris-babic-i-glenn-cohen-ariel-dora-stern-yiwen-li-me","title":"A general framework for governing marketed AI/ML medical devices","citingArticles":["ca-sb-243","hallucination"],"citeCount":2},{"key":"lit-emelie-stiernstr-mer-facial-recognition-technology-in","title":"Facial recognition technology in law enforcement: a scoping review of existing empirical studies","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-ezgi-yazici-toward-a-global-standard-for-ethical-ai-re","title":"Toward a global standard for ethical AI regulation: addressing gaps in AI-driven biometric and high-resolution satellite imaging in the EU AI Act","citingArticles":["fedramp-ai-guidance","gsa-ai-acquisition-guide"],"citeCount":2},{"key":"lit-hacker-engel-mauer-regulating-chatgpt-and-other-large","title":"Regulating ChatGPT and other Large Generative AI Models","citingArticles":["model-card","multi-turn-evaluation"],"citeCount":2},{"key":"lit-keith-raymond-harris-ai-or-your-lying-eyes-some-shortc","title":"AI or Your Lying Eyes: Some Shortcomings of Artificially Intelligent Deepfake Detectors","citingArticles":["china-deep-synthesis-provisions","provenance-watermarking"],"citeCount":2},{"key":"lit-lehdonvirta-w-hawkins-compute-north-vs-compute-south-t","title":"Compute North vs. Compute South: The Uneven Possibilities of Compute-based AI Governance Around the Globe","citingArticles":["designated-systemic","policy-instrument"],"citeCount":2},{"key":"lit-pengfei-li-jianyi-yang-mohammad-a-islam-shaolei-ren-ma","title":"Making AI Less 'Thirsty': Uncovering and Addressing the Secret Water Footprint of AI Models","citingArticles":["asean-ai-guide","oecd-ai-principles"],"citeCount":2},{"key":"lit-sandra-fredman-darcy-du-toit-alessio-bertolini-jonas-v","title":"Fair Work for Platform Workers: Lessons from the EU Directive and Beyond","citingArticles":["eu-platform-work-directive","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-shai-farber-machines-of-justice-a-systematic-review-of","title":"Machines of justice: A systematic review of AI applications in policing and criminal justice","citingArticles":["uk-ai-white-paper","unesco-ai-ethics-recommendation"],"citeCount":2},{"key":"lit-stephen-weymouth-digital-disintegration-techno-blocs-a","title":"Digital Disintegration: Techno-Blocs and Strategic Sovereignty in the AI Era","citingArticles":["asean-ai-guide","policy-instrument"],"citeCount":2},{"key":"lit-taner-kuru-lawfulness-of-the-mass-processing-of-public","title":"Lawfulness of the mass processing of publicly accessible online data to train large language models","citingArticles":["gdpr","retrieval-augmented-generation"],"citeCount":2},{"key":"lit-tobin-south-samuele-marro-thomas-hardjono-robert-mahar","title":"Authenticated Delegation and Authorized AI Agents","citingArticles":["eu-product-liability-directive","uk-us-aisi-mou"],"citeCount":2},{"key":"lit-vijaytha-muralidharan-madelena-y-ng-shada-alsalamah-sa","title":"Global Initiative on AI for Health (GI-AI4H): strategic priorities advancing governance across the United Nations","citingArticles":["ca-sb-243","hallucination"],"citeCount":2}]}]}