France's AI action summit put governance on the global table

For nearly a week, from February 6 to 11, 2025, Paris became the epicenter of global discussions on artificial intelligence governance. The AI Action Summit (Sommet pour l'action sur l'IA) packed the schedule with scientific conferences, a cultural weekend, an international summit at the Grand Palais, side events, and a dedicated business day. Its stated goal: strengthening international action for an AI serving the general interest, a principle echoed in the emerging global consensus on AI regulation.

The format was broad: research symposia alongside public exhibits. That wasn't an accident. The organizers wanted experts and citizens in the same conversation. The Grand Palais session brought together government delegations, industry leaders, and researchers. Side events and a business day were supposed to turn diplomatic talk into real projects. The involvement of major AI companies was notable, as seen in the joint policy letter signed by OpenAI, Meta, and 40 other organizations.

Paris wasn't a random choice. The city has been quietly building a reputation as a European AI hub, fueled by a wave of startups and government backing. Schools like École Polytechnique and CentraleSupécle churn out talent. Venture capital is flowing into deeptech. The summit adds a political layer to that groundwork. France wants a seat at the table where AI rules are written. Private sector investment in AI policy research signals a similar ambition, with companies like Anthropic doubling its commitment to public policy research.

What happens next is murkier. The summit produced statements of intent. Real binding rules? Not yet. For now, Paris hosted a good conversation. The next edition will test whether it can turn that talk into something concrete. The debate over open-weight models is a key part of that future, as argued by 41 companies in a recent open letter.