The Paris AI Action Summit: US and UK Refuse to Sign as Global AI Governance Splits

On February 10-11, 2025, the AI Action Summit co-hosted by France and India convened in Paris, where sixty-plus nations signed a declaration on open, inclusive and ethical AI — but the United States and United Kingdom refused, while China signed.

U.S. Vice President JD Vance used the stage to blast Europe's 'excessive regulation'; the U.K. cited national security and sovereignty. Per The Guardian, it marked the formal turn away from the Bletchley-to-Seoul 'AI safety summit' consensus track — safety receded, race narratives took over.

From Bletchley to Paris: A Consensus Track Ends in Three Years

The 2023 Bletchley summit had the U.S., China, U.K. and EU signing a safety declaration together; Seoul 2024 carried it forward. By Paris the summit's name had changed from 'Safety' to 'Action,' and the agenda from risk mitigation to investment and industry — the rename was itself the announcement of the pivot. Six months later the U.S. published its 'Winning the Race' action plan and deregulated across the board (see our coverage), cashing in the Paris signal in full.

The Geopolitics in the Signature List

The most telling detail is the alignment: China signed the open-and-inclusive declaration while the U.S. and U.K. refused — traditional allies openly parting ways on AI governance. The list foreshadowed the settled landscape: America on speed, the EU on risk (AI Act), China on content (labeling measures) — three regulatory paradigms each taking shape (see our policy coverage), with multilateralism reduced to a stage for blocs.

Our Take

Paris became the watershed of global AI governance: multilateral coordination gave way to bloc competition, and national regulatory paths diverged from there. The practical takeaway for operators: no globally unified AI rulebook is coming soon — going abroad means adapting compliance jurisdiction by jurisdiction, and the fragmentation of the regulatory map is itself a new cost of doing business.