U.S. Government Orders Anthropic to Shut Down Claude Fable 5 and Mythos 5 Models

The U.S. government on Friday ordered Anthropic to immediately shut off access to two of its most powerful AI models—Claude Fable 5 and Claude Mythos 5—citing national security concerns. Anthropic received the directive at 5:21 pm ET and has complied, disabling both models for all users worldwide. Access to Anthropic's other models remains unaffected.

The directive is framed as an export control action, but Anthropic says the underlying concern is a claimed jailbreak of Fable 5. According to the company, the government has provided only verbal evidence of a "potential narrow, non-universal jailbreak" that involves prompting the model to read specific codebases and identify software flaws.

The Models in Question

Claude Mythos 5 is Anthropic's most capable AI model, previewed in early April but kept tightly restricted due to its exceptional ability to find security vulnerabilities in software. According to Anthropic, Mythos identified flaws in every major operating system and web browser it tested. Rather than release it broadly, Anthropic launched Project Glasswing, sharing the model with approximately 50 vetted organizations including Amazon, Apple, Google, Microsoft, and CrowdStrike for defensive cybersecurity work.

Claude Fable 5, released three days ago, was Anthropic's commercial answer: a version of Mythos fitted with guardrails that block responses in high-risk areas like cybersecurity and biology. According to benchmark tests from Vals AI, it was immediately the most capable AI model available to the public.

Anthropic's Response

Anthropic disputes the government's decision in a lengthy blog post, arguing that the jailbreak represents a "level of capability" already widely available in other publicly accessible models, including OpenAI's GPT-5.5. The company notes that such capabilities are used routinely by cybersecurity professionals for defensive purposes.

The company also emphasizes that its strongest safeguards operate through independent classifier systems that function separately from the model itself. A review of recent usage found no evidence of those safeguards being successfully bypassed to produce harmful content.

"We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people," Anthropic wrote. "If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers."

Industry Implications

The shutdown creates an unusual situation for Anthropic, which has staked much of its public identity on being the safety-conscious alternative to rivals. The company's caution in restricting Mythos—promoted as too dangerous for public release—has now attracted government scrutiny that could disrupt its business ahead of a widely expected IPO this year.

OpenAI CEO Sam Altman previously criticized Anthropic's handling of Mythos in April, telling podcaster Ashlee Vance that it amounted to "fear-based marketing." Altman said: "It is clearly incredible marketing to say, 'We have built a bomb. We were about to drop it on your head. We will sell you a bomb shelter for $100 million.'"

What This Means

This marks the first time the U.S. government has ordered a complete shutdown of commercially deployed AI models over security concerns. The action sets a precedent for government intervention in AI deployments and raises questions about what constitutes an acceptable level of risk for public AI models. For Anthropic, the shutdown represents a significant setback: the company's safety-first positioning has paradoxically made it a target for exactly the kind of regulatory action it sought to avoid. The case also highlights the tension between AI companies' need to demonstrate capability to compete commercially while managing genuine safety risks—a balance that may become increasingly difficult as models grow more powerful.