US Government Ordered Anthropic to Pull Fable 5 and Mythos 5 Globally. They Did.

On June 9, 2026, Anthropic released Claude Fable 5 and Mythos 5 to the public. On June 12, 72 hours later, a directive from the US Commerce Department arrived. By the end of the evening, both models were gone for everyone.

Anthropic published a statement on its website confirming the suspension and explaining why it applied globally even though the order targeted a narrower group.

What the order said

The directive came from Commerce Secretary Howard Lutnick and was written with involvement from the Bureau of Industry and Security. The text, as Anthropic describes it, required suspending access to Fable 5 and Mythos 5 “by any foreign national, whether inside or outside the United States” including Anthropic’s own foreign national employees.

The catch: Anthropic has no reliable way to verify nationality in real time for every user on its platform. So to comply, it shut both models off for everyone. All other Claude models, including Opus 4.8, Sonnet 4.6, and Haiku, remain available.

Where the jailbreak came in

On June 10, a known jailbreaker who goes by “Pliny the Liberator” posted on X claiming to have bypassed Fable 5’s safety guardrails. The post claimed successful extraction of instructions related to cyber exploits, explosives, and chemical synthesis pathways. The post spread quickly.

Anthropic disputes the severity. The company says the identified issue is “narrow, non-universal” and that the vulnerabilities it surfaces are “already-known” and findable in other models. Its safeguards, Anthropic says, are “substantially more effective than those of any previously deployed model.”

The government letter did not, according to Anthropic, provide specific details about the national security concern it cited.

Anthropic’s position

The company is complying but not quietly. From the statement: “we disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model.”

Anthropic says it is “working to restore access as soon as possible” and is in contact with the relevant government bodies. Whether that means a patch, a negotiated resolution, or a longer standoff is unclear.

What this means for users

If you subscribed to a Fable 5-capable plan expecting to use those models, Anthropic opened a refund window. The window to request a refund or cancel runs into late June. Users who still want access to frontier Anthropic models can fall back to Opus 4.8, which remains fully available.

For developers building on the API, nothing changes for existing model IDs. Only claude-fable-5 and claude-mythos-5 are suspended.

This is the first time a major AI company has been compelled to take a commercial model offline under a government security directive. Whether the action holds, is challenged, or leads to broader export control frameworks for frontier models is now a question with real precedent.

---

Sources: Anthropic statement, CNBC, NBC News, Simon Willison’s summary