Anthropic has suspended access to its Fable 5 and Mythos 5 AI models after receiving a U.S. government export control directive citing national security concerns.

The company said it received the order on June 12 at 5:21 p.m. ET and complied by disabling the models for all customers worldwide, including foreign nationals employed by Anthropic. Access to the company’s other AI models remains unaffected.

In a statement, Anthropic said the government has not provided detailed information about the alleged security risk. The company believes the directive stems from concerns over a reported “jailbreak” technique that could bypass Fable 5’s safeguards.

Anthropic said its review found the technique exposed only a small number of previously known, minor vulnerabilities and did not demonstrate capabilities beyond those available in other publicly accessible AI models.

The San Francisco-based startup defended Fable 5’s security measures, noting that the model underwent extensive red-teaming by government agencies, independent organizations, and internal teams before launch. The company said no universal jailbreak capable of broadly circumventing safeguards has been identified.