OpenAI has announced AI Futures, a new Strategic Futures team launched today to research how democratic societies should be restructured to preserve individual rights and agency amid transformative AI development. Led by Dean Ball, the team frames the core challenge as one of concentration of power: as autonomous systems and machine intelligence advance, states may no longer need widespread human cooperation to project force or generate revenue, potentially enabling governance without societal consent. The team will publish research at the intersection of public policy, economics, law, history, and machine learning via blog posts, papers, videos, and podcasts.
The Strategic Futures initiative centers on a structural worry that transcends current partisan politics: as AI automates bureaucracy, military function, and tax collection, the traditional bargain between rulers and ruled—historically mediated through labor and consent—may dissolve. OpenAI frames this as the most serious long-term AI risk category and argues that avoiding concentration of power requires balancing incentives and checks between actors, drawing on Madisonian federalist principles applied to the AI era rather than radical decentralization.
For policy architects, AI labs, and governance teams, this signals OpenAI's investment in foundational research on AI's political economy and institutional design, separate from product safety or capabilities. The team will explore how firms, government agencies, and civil society might restructure themselves as AI reshapes labor markets and state capacity. This echoes similar moves by other frontier labs (Anthropic's policy team, DeepMind's governance research) and suggests that concentration-of-power questions are now table stakes for public legitimacy of frontier AI development.