Eight claims the working group presupposes rather than proves — the internal logic from which the Provocations depart.
Information is the produced, stored, circulated, and now algorithmically generated material through which a polity perceives itself, deliberates, and acts. The systems that filter, rank, and generate information are constitutive. The object of democratic theory shifts accordingly: from the exchange between already-formed subjects to the infrastructure that forms the conditions of knowing, deciding, and acting.
Democracy's present crises are best diagnosed not as failures of conversation but as failures in the infrastructures through which societies produce, distribute, and contest information. The question is no longer how to improve deliberation within existing arrangements, but how those arrangements are built, owned, and governed in the first place.
The same infrastructures that could deepen democracy are currently corroding it. The question is therefore not whether to build them, but who governs them and to what end. Treating technology as a neutral instrument concedes the most important decisions before they are made.
Democratic institutions can be deliberately redesigned as informational regimes — as systems for collecting knowledge at scale, deliberating across difference, and delivering results. The absence of a field equipped to do this work is itself part of the problem this group sets out to address.
The deeper problem is not technology as such but the absence of infrastructures through which societies can decide what may be trusted, how claims can be contested, and how collective sense-making can happen under complexity. The same systems that could deepen democracy are currently corroding it. The question is therefore not whether to build them, but who governs them, by what legitimacy, and to what end.
AI systems no longer merely move content through existing channels; they perform epistemic functions — searching, ranking, summarizing, recommending, advising, moderating, acting. The stake is agency: whether publics and institutions retain the capacity to know and decide for themselves, or cede it by default.
The promise of AI is not that machines replace citizens, experts, or institutions, but that democratic societies might become more capable of sensing problems, synthesizing knowledge, representing plural perspectives, and learning from consequences over time. The design challenge is to build infrastructures that make AI augment democratic intelligence rather than concentrate cognitive power or substitute for collective judgement.
Substantial work already exists — on the public sphere, deliberative and epistemic democracy, platform governance, collective intelligence, AI governance, and civic technology — but it remains fragmented across disciplines. What is missing is an integrated, design-oriented field that asks how democracies can deliberately build the conditions for trust, knowledge, and agency in AI-mediated societies.