Key Points
- A US Commerce Department order barred foreign access to Claude Fable 5 and Mythos 5 on June 12.
- Anthropic disabled both models for all customers to ensure compliance; its other models are unaffected.
- The order covers any foreign national inside or outside the US, including Anthropic's foreign staff.
- Commerce acted after another company claimed it jailbroke one of the models, citing national security.
Washington just treated two specific AI models like controlled munitions. On June 12, the US Commerce Department ordered Anthropic to cut off foreign access to its two most advanced models — and rather than build a carve-out on the fly, the company pulled them for everyone.
What happened
According to the order, Commerce Secretary Howard Lutnick directed Anthropic CEO Dario Amodei to suspend all access to Claude Fable 5 and Mythos 5 "by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees." Anthropic says it received the directive at 5:21 p.m. ET and disabled both models for all customers to ensure compliance. The company's other models, it said, continue to operate normally.
Why the order came
An administration official told Axios that Commerce acted after another company claimed it had jailbroken the models, raising national-security concerns; Anthropic's own statement describes the issue as a narrow potential jailbreak of Fable 5 — asking the model to read a codebase and fix software flaws. Anthropic has pushed back, arguing a narrow potential jailbreak should not be cause for recalling a model deployed to hundreds of millions of people, and describing the episode as a misunderstanding it is working to resolve so it can restore access.
The read
Export controls have governed advanced chips for years — the rules that keep Nvidia's top GPUs out of China. Pointing them at a single frontier model, and at foreign persons inside the US, is an escalation: it treats model access itself as a national-security asset, not just the hardware underneath it. The timing sharpens it. Anthropic is days into a confidential IPO process, and its most valuable product is now subject to direct federal on-off control. Whether this resolves in days or sets precedent, every frontier lab just watched the government reach into a live product.
That posture has kept Anthropic outside the industry's open-weights bloc — it did not sign Jensen Huang's July 24 open-weights letter and is absent from Nvidia's Open Secure AI Alliance.
Comments