Anthropic received a directive from the U.S. government ordering it to suspend access to its Fable 5 and Mythos 5 models for all foreign nationals, including foreign national employees, citing national security export control authorities. The net effect forces Anthropic to disable both models for all customers to ensure compliance. Access to other Anthropic models is still good to go
A sudden directive without details
So far, the government’s letter has provided no specific details of its national security concern. Anthropic’s understanding is that officials believe they’ve discovered a method of “jailbreaking” Fable 5, bypassing its safeguards. However, when Anthropic reviewed the demonstration, it found the technique identified “a small number of previously known, minor vulnerabilities” that are “relatively simple” and that “other publicly-available models can discover them as well without requiring a bypass.”
Anthropic had already publicly stated when launching Fable 5 that perfect jailbreak resistance is not currently possible for any model provider. The company adopted a “defense in depth” strategy: making jailbreaks narrow or expensive to produce, combined with thorough monitoring to detect and shut down attacks. Anthropic also required 30-day customer data retention specifically to research and mitigate jailbreaks.
To date, the government has only provided verbal evidence of what Anthropic calls a “narrow, non-universal jailbreak,” essentially asking the model to read a codebase and fix software flaws. Anthropic validated that the level of capability displayed “is widely available from other models (including OpenAI’s GPT-5.5) and is used every day by the defenders who keep systems safe.”
The Pentagon connection: Weaning off Anthropic
The Fable 5 suspension arrives amid broader U.S. government distancing from Anthropic. The Pentagon has switched at least two-thirds of its use of Anthropic to competing artificial intelligence (AI) providers, according to Emil Michael, the Department of Defense’s chief technology officer.
The comments came during a six-month wind-down in the use of Anthropic’s tools after the Defense Department designated the firm as a supply chain risk amid a contracting fight. While the directive on Fable 5 came from export control authorities, not the Pentagon itself, the vibe is pretty obvious: the government is actively trying to break up with Anthropic.
The company has not been told the specific “national security concern,” but the cumulative effect (a sudden export control order and a Pentagon divestment) suggests the national security crowd is getting seriously nervous about Anthropic’s tech or how they handle security.
What Anthropic is saying: Disagreement and compliance
Anthropic’s statement is notable for its frankness. The company is complying with the legal directive but “disagrees that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people.”
Anthropic argues that if this standard were applied across the industry, “it would essentially halt all new model deployments for all frontier model providers.” The company has long advocated for the government to have the ability to block unsafe deployments, but through “a statutory process that is transparent, fair, clear, and grounded in technical facts.”
This action, Anthropic says, “does not adhere to those principles.” The company apologized to customers for the disruption, called it a “misunderstanding,” and said it is “working to restore access as soon as possible.”
