Open problem 13
The Democratic Capacity Problem
- current AI
- frontier AI
- AGI
How can democratic institutions govern AI when AI systems may themselves reshape information, persuasion, administration, expertise, lobbying, surveillance, and political mobilisation?
Why it’s foundational
AI governance assumes a functioning public sphere, capable state institutions, and meaningful accountability. But AI may alter all three.
Why it’s difficult
The effects are diffuse and endogenous. AI can improve public services and citizen access to knowledge, while also enabling personalised propaganda, synthetic media, automated lobbying, bureaucratic opacity, and dependence on private infrastructure.
Hidden assumptions
A major hidden assumption is that democratic institutions are external to AI and can regulate it from outside. Increasingly, governments will use AI internally, political campaigns will use AI strategically, and citizens will encounter politics through AI-mediated channels.
Competing positions
- Digital-democracy optimism
- AI-as-propaganda-risk
- Administrative-state modernisation
- Surveillance-state critique
- Platform-governance approaches
- Public AI infrastructure
What could make progress
Field experiments on AI persuasion; studies of AI use in public administration; transparency requirements for political AI; audits of automated lobbying; research on AI-mediated news consumption; institutional design for public-interest AI systems. AISI’s Frontier AI Trends Report already treats persuasion, political information-seeking, emotional dependence, and high-stakes agent use as emerging societal-impact areas; the deeper research problem is how those capabilities affect democratic authority itself.
What it would change
It would affect election law, campaign regulation, public-sector AI procurement, transparency duties, media policy, and democratic oversight of state AI.
Sub-agenda
- When does AI persuasion become democratically illegitimate?
- How should law treat personalised political AI agents?
- What public-sector AI uses require heightened democratic scrutiny?
- Can AI improve deliberation without manipulating participants?
- Should democracies build public AI infrastructure rather than rely on private systems?
Priority (editor scoring)
Underexplored as a governance-capacity problem rather than a misinformation topic.
- Importance
- 4/5
- Neglected
- 4/5
- Difficulty
- 4/5
- Actionable
- 4/5
- Robust
- 5/5
- Nat’l+int’l
- 3/5
Where the catalog bears on this
No current catalog instrument resolves this puzzle — which is the point: it is a foundational question the existing rules leave open. Browse the coverage catalog for what the instruments do and don’t say.
Editorial content — a human-authored agenda question, rendered verbatim. No part of this analysis is AI-generated (see the charter).