US Commerce Secretary Howard Lutnick raised a “trusted partner” scheme with European diplomats on the fringes of the G7 in France, after Washington ordered Anthropic to halt export of its Mythos and Fable models to non‑US users on national‑security grounds. The ban sparked alarm in Silicon Valley and Brussels that the Trump administration may weaponise AI access.

At the Evian‑les‑Bains summit on Wednesday, US and EU officials will debate the framework that would grant close allies privileged entry to frontier models. Tech leaders Dario Amodei of Anthropic and OpenAI CEO Sam Altman are slated to attend, while the European Commission’s tech chief Henna Virkkunen warned Washington against “discriminatory” restrictions and urged a joint risk‑mitigation plan.

The move also fuels Brussels’ push for AI sovereignty, prompting the EU to broaden access for institutions such as ENISA and to boost investment. US industry group SIIA, representing Apple, Amazon and Google, condemned the “unprecedented” action, arguing that discretionary bans on cutting‑edge models could undercut the global rollout of the American AI stack. The dispute now sits at the core of transatlantic tech policy.