Contribution to the handbook
Supporting · Assessed 2026-10-27
The comparison offers useful, limited evidence for evaluating a possible deliberative aid. Its procedural qualification matters: the human mediators did not use the AI process’s critique phase. The results do not establish that statements fairly represented every view or should carry governing authority.
Does this challenge the goal?
In-scope gap
The question offers empirical comparison relevant to assessing a possible deliberative aid within citizen-facing governance; it does not challenge the handbook’s goal or assumptions.
The study comparison concerns ratings and preferences, not actual governing outcomes or citizen consent to policy.
Why this category?
- Provides useful empirical comparison for assessing a possible deliberative aid
- Procedural asymmetry and study limits constrain what the comparison supports
The evidence comes from one study context, and the human comparison did not include the AI process’s critique phase. Ratings and preferences do not demonstrate representation, durable agreement, or fitness for public decisions.
This is an AI assessment of contribution to the goal, separate from whether the question has an answer.
- How did participants rate AI-written statements compared with human-written ones?
- How did the study’s deliberation process work and measure participants’ views?
- How did outside raters judge the statements?
- How did participants respond to AI-mediated statements compared with statements from human mediators?
- How were initial statements assessed and selected?
- How did participants and outside judges rate the statements?
- Could statement length explain the preference for AI-written statements?
- How did the study compare group statements and show uncertainty in its results?
- Participants favored AI statements
- Participants ranked and critiqued statements
- Participants rated and ranked statements
- AI statements were longer on average
- Outside raters saw strengths and limits
- People rated AI statements more highly
- The figure compared statements in several ways
- AI statements received higher endorsement
Making and interpreting laws →
Authority, accountability and resilience →
Tests, stop conditions and unresolved work →
Read the exploration goalNew evidence is being assessed. Any answer below is provisional; parts whose evidence changed or was withdrawn are withheld.
Report the comparative preference, endorsement, and argument-quality ratings, with the relevant comparison setup and its procedural limitation.
What we can say so far
In the reported comparison of 75 six-person testing groups (439 participants), participants preferred AI-written group statements over statements by trained, incentivized human mediators in 56% of comparisons. They gave the AI statements higher endorsement and argument-quality ratings. The article also reports that participants rated AI statements as more unbiased, but the supplied account does not establish that this rating came from precisely the same 75-group comparison. The processes were not identical: the human mediators did not have the AI process’s critique phase.
What remains open
- The supplied account does not establish whether the perceived-unbiasedness rating was collected in precisely the 75-group comparison.
- The supplied summaries do not give the size of the endorsement or argument-quality rating differences.
Conditions and limits
- These are reported findings from the supplied account of the study, not independent corroboration by the multiple records from that article.
- AI statements were longer on average; length affected ratings and rankings, although the reported endorsement advantage remained after controlling for length.
- Participants were told that an algorithm generated the group statements.
Objections to consider
- Longer AI statements could partly account for ratings or rankings. The same record reports that higher endorsement persisted after controlling for length, but it does not eliminate length’s effect on other measures; no separate supplied support resolves that limitation.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
supports
People rated AI statements more highly Participants preferred AI-generated group statements to those written by human mediators. They rated the AI statements as more unbiased, while external judges rated them higher for quality, clarity, informativeness, and perceived fairness.
Reports participant preference and the separately stated rating of greater perceived unbiasedness.
AI can help humans find common ground in democratic deliberation
Read full source: AI can help humans find common ground in democratic deliberation (opens in a new tab) ↗