G7 leaders are negotiating with the U.S. to give select 'trusted partners'—countries or companies—access to advanced AI models from firms like Anthropic, after the Trump administration ordered restrictions on foreign access for national security reasons.

What happened: Anthropic disabled access to its most advanced AI models, Fable 5 and Mythos 5, for all users after President Trump ordered the company to block foreign nationals from accessing them. G7 leaders discussed with U.S. Commerce Secretary Howard Lutnick at the summit in Evian-les-Bains, France, the possibility of granting select trusted partners broader access to advanced AI models from U.S. firms.

Why it matters: A Trump administration official confirmed the U.S. team has 'an open line of communication with our allies' and remains committed to addressing national security concerns. If an agreement is reached, G7 countries could use advanced models to develop stronger cybersecurity defenses against rivals such as China. Anthropic's Mythos model, which is designed to find flaws in computer code, is seen by cybersecurity experts as potentially able to turbocharge attacks on banks' technology systems, which is why the European Union is seeking access to study it.

What to watch: AI executives from Anthropic, OpenAI, and Google are expected to attend a working lunch on Wednesday to discuss technology issues, including regulation, AI infrastructure, and networks. Prior to Trump's order, Anthropic had given access to Mythos to select organizations in 'more than 15 countries' in the healthcare, communications, power, and water sectors.