Question the Questions: Fairer Expert Q&As in Citizens’ Deliberations
In citizens’ assemblies and online town halls, many people submit questions for experts—but only a few make it on stage. How do we choose a small set that truly reflects everyone’s concerns?
Researchers introduce a way to audit representation of the chosen questions, grounded in a social-choice idea called justified representation. Their algorithms check whether key participant groups are fairly covered and spotlight what’s missing.
Applied to real deliberations, they compared:
- Questions picked by a moderator
- Questions optimized from participants’ inputs
- Summary questions generated by large language models (LLMs)
Results: LLMs show promise but can still overlook some voices. Auditing helps organizers fix gaps before the session.
Best of all, these tools are integrated into an online platform used in hundreds of deliberations across 50+ countries—making fairer Q&A practical, not just theoretical.
Paper: http://arxiv.org/abs/2511.04588v1
Paper: http://arxiv.org/abs/2511.04588v1
Register: https://www.AiFeta.com
civictech deliberation publicpolicy AI LLM socialchoice representation democracy
