The US government just ordered Anthropic to shut down its two most powerful AI models - for everyone - because of national security concerns. On Friday evening, the Trump Administration's Commerce Department issued an export control directive blocking all foreign nationals from accessing Fable 5 and Mythos 5, Anthropic's most capable AI systems. Because complying with that order while keeping the models running for everyone else was impossible to guarantee, Anthropic had to disable both models for all users worldwide. The directive arrived at 5:21pm on a Friday. No specific details were provided. Anthropic learned about it the same time everyone else did. The government's concern centres on Mythos - an AI system so capable at detecting software vulnerabilities that it has been used by US authorities to find and fix security gaps that had gone unnoticed for decades. The fear is that the same capability in the wrong hands becomes a cyberweapon. Anthropic disputes the reasoning. The company says it reviewed the jailbreak technique that triggered the order and found the same capability is already available in OpenAI's GPT-5.5 and other public models. It also warned that if this standard were applied industry-wide, it would effectively halt all new frontier AI model deployments across every company. This is the furthest-reaching government action ever taken against a commercial AI model.