Cross-corpus research synthesis
Compute + Model-Weight Export Controls
Restrictions on cross-border flow of frontier AI compute (GPUs, accelerators) and model weights. Distinct from `compute_reporting` (which is disclosure) — this is restriction of access by recipient. US BIS rules (Oct 2023 advanced computing, Jan 2025 outbound investment), EU dual-use Regulation 2021/821 overlay, China retaliatory measures + indigenisation push. Mostly outside traditional AI-governance instruments; carving its own track.
Synthesised deterministically from 6 articles that engage this theme. Empirical consensus: contested · contested: 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.. Full theme article: /wiki/compute-export-controls. Machine-readable: /wiki/synthesis.json.
Cross-jurisdiction stances (0 govern, 4 engage)
| Instrument | Verdict | Provision excerpt / citation |
|---|---|---|
| Executive Order 14110 on Safe, Secure, Trustworthy AI | implicit | §4.2(b) directs export-control coordination via BIS; not the primary venue but the policy hook |
| Anthropic Responsible Scaling Policy (RSP) v2 | implicit | ASL-3+ tiers include model-weight access controls (recipient-restriction analog) |
| Google DeepMind Frontier Safety Framework | implicit | FSF mitigations include model-weight access controls + restricted-deployment options |
| Meta Frontier AI Framework | implicit | Framework's release decisions implicitly determine cross-border weight flow |
Evidence convergence
Sources the corpus cites for this theme across multiple articles — a scientometric consensus signal computed from inline prose citations (the more articles independently cite a source, the more load-bearing it is for this theme). 14 sources are cited by ≥2 articles.
- 5×China's semiconductor conundrum: understanding US export controls and their efficacy — cited by 5 articles
- 4×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 — cited by 4 articles
- 4×Artificial intelligence and synthetic biology: biosecurity risks, dual-use concerns, and governance pathways — cited by 4 articles
- 3×Infrastructure for AI Agents — cited by 3 articles
- 3×Two types of AI existential risk: decisive and accumulative — cited by 3 articles
- 3×Multi-Agent Risks from Advanced AI — cited by 3 articles
- 3×The EU model of AI governance: regulating artificial intelligence through law and policy — cited by 3 articles
- 3×Governing AI Agents — cited by 3 articles
- 2×Missing the Mark: Adoption of Watermarking for Generative AI Systems in Practice and Implications Under the New EU AI Act — cited by 2 articles
- 2×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 — cited by 2 articles
- 2×Training Compute Thresholds: Features and Functions in AI Regulation — cited by 2 articles
- 2×Defending Compute Thresholds Against Legal Loopholes — cited by 2 articles
- 2×International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty — cited by 2 articles
- 2×Computing Power and the Governance of Artificial Intelligence — cited by 2 articles