Anthropic disabled its newest AI models, Fable 5 and Mythos 5, after the U.S. Commerce Department used national security export controls to bar the company from distributing the models to foreign nationals. The directive extended to foreign nationals located in the U.S., including Anthropic non-citizen employees. Anthropic said it disabled access for all users because of the scope of the directive and that access to less powerful Claude models, including Claude Opus 4.8, was not affected. The company also said it received notice at 5:21 p.m. Eastern Time without specific details, but was told the action followed a concern about a technique that could bypass safeguards. The dispute has immediate downstream implications for universities and research labs that rely on commercial model access for pilots and coursework, particularly where policies constrain user eligibility by nationality. It also further sharpens the compliance burden for vendors distributing models across borders under national security frameworks.