Top AI CEOs push US-led global standards at G7 summit

The leaders of Anthropic and Google DeepMind used a closed-door G7 luncheon in France to urge the Trump administration to spearhead international rules for artificial intelligence, according to sources familiar with the meeting.

Anthropic CEO Dario Amodei and DeepMind chief Demis Hassabis proposed that the United States take the lead in coordinating global AI governance, citing the need to manage risks from increasingly capable models. Canadian Prime Minister Mark Carney reportedly backed the idea of a US-led coalition.

The session, held Wednesday in Évian-les-Bains, included about a dozen tech executives and G7 heads of state. OpenAI's Sam Altman also attended, calling for an international forum to establish shared testing standards and provide impartial risk analysis.

Amodei outlined specific areas for cooperation: structured access to frontier models, chip trade policies that exclude China, and joint action on AI threats related to cybersecurity, bioterrorism, and intelligence. His comments come days after the US government imposed export controls on Anthropic's newest models, Fable 5 and Mythos 5, citing national security.

OpenAI's global affairs chief Chris Lehane noted that non-US leaders in the room acknowledged Washington's natural role in setting AI standards. The push for coordinated governance reflects growing industry concern that advanced cyber capabilities could cause major harm if misused.

Source: CNBC