Contribution to the handbook
Priority · Assessed 2026-10-28
Public willingness to delegate influence, the distinction between influence and final authority, and the practical limits of human review are consequential to deciding what role AI could have in governance. The evidence preserves competing considerations: some reported openness to limited influence, continued preference for experts, and concerns about scrutiny and reviewability.
Does this challenge the goal?
In-scope gap
The question examines public willingness and the boundary between AI influence and decision authority, both open questions in the existing exploration; it does not itself contest the goal.
The supplied public-attitude findings are limited and do not establish support for AI having final authority.
Why this category?
- Addresses a consequential open choice about AI influence and human authority
- Includes relevant public-attitude evidence with limits on its scope
- Identifies reviewability and effective human control as material risks
The reported willingness figures come from cited research whose sample and task details are limited in this packet. Stated willingness is not actual use or endorsement of AI control; formal monitoring or override also does not establish that review is effective.
This is an AI assessment of contribution to the goal, separate from whether the question has an answer.
- Do people want AI to make technical decisions?
- Three levels of human control
- People need to examine the factors
- Deference depends on the outcome
- Some systems resist review
- Acceptance of AI in politics is unsettled
- Experts remain preferred
- Concern and openness can coexist
- Some support for limited AI influence
Making and interpreting laws →
Judging situations and resolving disputes →
Authority, accountability and resilience →
Read the exploration goalAssess evidence of public willingness to let AI make technical decisions, distinguishing AI influence from final decision-making authority.
What we can say so far
The supplied evidence indicates some willingness to give AI influence in technical decisions, but also reports a preference for experts. It therefore supports limited openness to AI influence, not a conclusion that people want AI to have final decision-making authority. The evidence does not establish public willingness to give AI final authority over technical decisions.
What remains open
- Evidence specifically measuring willingness to let AI make final technical decisions is not supplied.
- The task definitions, study sample and other methodological details needed to assess how broadly the reported finding applies are not provided.
Conditions and limits
- The technical-task finding is reported by the authors from a cited study, rather than from their own experiment.
- The supplied passage does not specify which technical tasks were considered.
- The reported willingness concerns AI influence; experts remained preferred.
- The cited 2024 result concerns a decision process in which AI had 25% influence and politicians 75%, not AI final authority or necessarily technical decisions.
- The evidence reports stated willingness, not actual behavior.
Objections to consider
- The reported 25% AI-influence result may appear to support public acceptance, but its political setting and limited AI role do not answer whether people want AI to make technical decisions or hold final authority.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
supports
Experts remain preferred The same cited study found that people were more willing to give AI some influence on technical tasks, but still preferred to defer to experts.
Directly reports greater willingness to give AI some influence on technical tasks while experts remain preferred, supporting a qualified distinction between influence and final authority.
Large Language Models (LLMs) as Agents for Augmented Democracy
Read full source: Large Language Models (LLMs) as Agents for Augmented Democracy (opens in a new tab) ↗