Anthropic chief executive Dario Amodei warned G7 leaders in Évian‑les‑Bains that democratic nations must avoid “splintering” on advanced AI deployment after Washington barred his firm from exporting its newest models. Speaking alongside US President Donald Trump, Amodei urged a unified stance to keep powerful tools from falling into hostile hands.

OpenAI’s Sam Altman and DeepMind founder Demis Hassabis joined Amodei, calling for US‑led cooperation on model safety, cyber‑defence and bioterrorism risks. The block targeted Anthropic’s Mythos and Fable systems, which officials said could bypass safety guardrails. European firms fear losing access to cutting‑edge AI, while US tech companies worry about revenue loss from restricted exports.

French President Emmanuel Macron said the dispute “clarified the stakes” for the US and its allies, urging stronger AI regulation and a shared standards platform. Indian Prime Minister Narendra Modi echoed the call for democratic access to frontier models. Facing supply‑chain worries, the G7 must lock in a framework rather than let rival blocs fragment the market. Time is running out for decisive action.