Contribution to the handbook
Priority · Assessed 2026-10-27
The comparison materially qualifies what the assembly results can support: the test mainly reproduced discussion and omitted fact-finding and expert testimony. The reported small-group testing and limits on generalization also matter when judging whether these results inform implementation or further tests.
Does this challenge the goal?
In-scope gap
Comparing what a test reproduced and omitted is within the definition’s emphasis on testing and on limits to implementation; it does not contest the goal.
The question concerns one virtual assembly, not whether AI should govern.
Why this category?
- consequential limits
- testing
- implementation
These records describe one study’s design and outcomes. They do not show whether adding the omitted stages would change results, or whether findings generalize to other populations, larger groups, or governance settings.
This is an AI assessment of contribution to the goal, separate from whether the question has an answer.
Authority, accountability and resilience →
Implementation and transition →
Tests, stop conditions and unresolved work →
Read the exploration goalCompare the stages the virtual assembly reproduced with those of a full citizens’ assembly that it omitted.
What we can say so far
The virtual assembly reproduced the discussion/deliberation phase, rather than a full citizens’ assembly. Participants used the study’s deliberation procedure: they provided views, considered AI-generated group statements, ranked them, critiqued a selected statement, and considered revised statements. In the assembly, groups discussed nine shared questions over three weekly one-hour sessions. The article says the assembly did not include the usual fact-gathering or expert-testimony stages. Thus it matched a defined discussion component, but omitted preparatory and informational stages of a full assembly; the supplied evidence does not establish that the two formats are equivalent overall.
What remains open
- The evidence does not report whether adding fact-finding or expert testimony would have changed the deliberation or its results.
- It does not establish equivalence between this virtual design and a full citizens’ assembly beyond the discussion-phase comparison.
Conditions and limits
- This is a comparison of the reported study design, not an independent assessment of how closely the experience or outcomes matched a full citizens’ assembly.
- The assembly’s reported schedule and nine shared questions are specific to this study; the evidence does not establish that the design represents every citizens’ assembly.
- The source is the supplied interpretation of the article; supplementary materials are not included.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
supports
The AI revised statements using feedback The AI helper drafted a group statement from participants’ opinions, then revised it using their critiques. The design drew on Jürgen Habermas’s theory of communicative action; across successive rounds, researchers used supervised fine-tuning and reward modeling to improve the system, with group approval as the stated goal.
Describes the statement-drafting and revision process that helps characterize the reproduced deliberation procedure.
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) ↗