- Possible justification
- The human role at stake is directing one’s own life, with the support one wants. That can matter even when AI performs a task better. Assistance may enable wanted communication, relationships or participation; it should not impose unwanted supervision. Building on the linked Human Rights Watch and CIEDUR research, HBC tests these choices as AI advances.
- Strongest objection considered
- A funded human option can still leave people with little say, few providers or unsuitable schedules. A family member may value relief that the recipient experiences as lost control. Protecting workers can conflict with an immediate request to change assistants; unrestricted employer power can expose assistants to insecurity or abuse. Neither a staffing label nor collective ownership settles these conflicts.
- Who might benefit or be excluded
- Examine the person receiving support, the assistant, relatives, excluded applicants and people without family advocates separately. Do not assume a disabled person is only a recipient: they may also be a parent, worker or carer. A service user’s communication requirements should shape how their preferences are elicited. More paid support need not mean more control; less family work need not mean the recipient chose the replacement.
- Resources and costs to investigate
- Cost suitable assistance, cover during absence, accessible communication, training, administration and effective challenges. Ask who pays for recruitment, changes and continuity, and who absorbs unpaid work when provision fails. Compare funding adequate for wanted support with competing needs and public resources. No net saving or sufficient budget is established here.
- An alternative to compare
- Compare a funded human option with capable AI, assistive devices, accessible environments, other providers and combinations the person can actually use. Technology may increase privacy or reduce unwanted dependence on another person. Compare safety, usability, control, cost and the ability to change course; do not make human contact compulsory or treat a cheaper tool as an adequate replacement without testing the person’s wanted outcome.
- What to check before using this elsewhere
- A proposal needs a responsible funding authority, a lawful eligibility and allocation route, accessible support for expressing preferences, and a workable way to choose, change or challenge provision. Protect assistants’ rights and plan continuity when a relationship ends. Specify resources and feasible alternatives for people outside the scheme. A person’s representative, provider and household should not be presumed to have identical interests.
- Evidence that could change the assessment
- The case would strengthen if people, including those needing communication support and those usually excluded, obtained more usable control and wanted support with fair working conditions and acceptable costs. It would weaken if the design mainly entrenched provider or family control, displaced support people preferred, shifted unwanted work to others or performed worse than a feasible accessible alternative.
These are HBC’s judgments and questions for further research. They do not show actual benefits or costs, establish workers’ preferences or authorise us to speak for affected people.