Anthropic disables Fable 5, Mythos 5 after US directive citing 'national security concern'

Anthropic said Friday it is disabling its "Fable 5" and "Mythos 5" models for all users after the US government issued an export control directive.

"The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees," said the company, adding that access to all of its other AI models will not be affected.

Anthropic said it received the order at 5.21 pm ET (2121GMT) Friday and that the letter did not explain the government's specific security concern in detail.

According to the company, the government believes someone has found a way to "jailbreak," or bypass the safety controls of Fable 5. Anthropic said the issues shown to it were already known, were "minor," and could also be found by other publicly available AI models without needing any special bypass.

The company argued that before releasing Fable 5, it spent thousands of hours testing the model's safety systems together with the US government, the UK's AI safety body and other groups, and that no one has found a way to fully bypass its protections.

"We suspect that perfect jailbreak resistance is not currently possible for any model provider," said the company, noting that all AI safety systems in the industry have some weaknesses, and that fully effective bypass methods will likely be discovered for any model eventually.

Anthropic said the government has only given verbal details of one limited bypass method, which it described as "asking the model to read a specific codebase and fix any software flaws."

It also noted that it reviewed a report it believes is behind the government's order and found that the same capability is "widely available from other models (including OpenAI's GPT-5.5), and is used every day by the defenders who keep systems safe."

While saying it will follow the government's order, Anthropic said it disagrees that the kind of limited issue should lead to pulling back a commercial AI model used by hundreds of millions of people.

It said applying this standard across the industry "would essentially halt all new model deployments for all frontier model providers."

Saying that it believes the situation is the result of "a misunderstanding," the company assured users that it is working to restore access to the models as soon as possible.