Anthropic announced it must disable Fable 5 and Mythos 5 for all customers following a US government export control directive issued at 5:21pm ET on national security grounds. The order bars access by any foreign national, including the company's own international employees, effective immediately.

The government expressed concerns about a potential jailbreaking method that could bypass safeguards, though Anthropic found the demonstrated vulnerabilities to be minor and widely reproducible across other publicly available models. The company maintains its defense-in-depth approach includes safeguards stronger than any previously deployed model.

Anthropic conducted extensive red-teaming with the US government, UK AISI, and private organizations for thousands of hours before Fable 5's launch. No universal jailbreak was discovered during testing, and the company argues that perfect jailbreak resistance is impossible for any model provider.

The single disclosed potential jailbreak involves asking the model to read code and fix software flaws—a capability available in models like GPT-5.5. Anthropic believes recalling commercial models for narrow jailbreak findings would halt all frontier AI deployments industry-wide, and disagrees with the government's handling of this matter.