Trump Officials Deny G7 Allies Access to Anthropic's Most Advanced AI

Allies Locked Out of Frontier AI

The chief executives of the world's most powerful artificial intelligence companies gathered alongside heads of state at the G7 summit in Évian-les-Bains, France, where AI governance emerged as the central issue. The talks were complicated by a fresh U.S. export control order that has cut off allied nations from one of the industry's most advanced systems.

Days before leaders convened, the Trump administration imposed export controls on Anthropic's most advanced AI models, Fable 5 and Mythos 5, ordering the company to halt access for all foreign nationals, including those inside the United States. Commerce Secretary Howard Lutnick cited concerns that the models could be exploited by military intelligence agencies in China, Russia, or other adversarial nations. Because Anthropic could not reliably verify users' nationalities, it disabled both models entirely.

A First Under the 2018 Export Reform Act

The New York Post reported that Trump officials have declined to grant G7 allies exemptions from the restrictions. The move represents the first time the Commerce Department has used its authority under the 2018 Export Control Reform Act to regulate AI as an emerging technology. Allied governments found themselves locked out of frontier AI tools developed in America, even as Washington urged collective action on AI governance and supply chain security.

The Summit Agenda

OpenAI's Sam Altman, Anthropic's Dario Amodei, and Google DeepMind's Demis Hassabis all attended the summit, which ran from 15 to 17 June. Mistral AI's Arthur Mensch was also among the attendees. French President Emmanuel Macron personally invited Altman to the Leaders Summit, marking the OpenAI chief's first appearance at such an event.

The summit's digital agenda centred on safe AI deployment, online child protection, digital security infrastructure, and securing critical mineral supply chains. Ahead of the gathering, OpenAI called for the creation of an international youth AI safety institute to develop guidance and support stronger safety standards across countries. The summit was expected to produce voluntary pledges rather than binding commitments.

Backlash From Allies

The White House framed U.S. participation around promoting American AI leadership and securing mineral supply chains, a focal point given China's dominance over resources essential for AI hardware, electric vehicles, and defence technology. The export order triggered a sharp backlash, with France announcing a break from American AI tools, reportedly dropping Palantir from a spy agency, and Canada warning of systemic risks from overreliance on US providers. Japan's Prime Minister Sanae Takaichi proposed a minerals initiative as G7 nations sought to reduce reliance on China.