June 12, 2026
The US government ordered Anthropic to suspend worldwide access to two of its newest models, Claude Fable 5 and Claude Mythos 5, for all foreign nationals, and the company disabled them for every customer to comply. The Commerce Department lifted the order 18 days later, on June 30, and Anthropic restored the models.
Anthropic said the directive, issued under national-security and export-control authorities, arrived on the afternoon of June 12 and required it to block access for every foreign national, inside or outside the United States, including its own foreign-national employees. Because it could not enforce that selectively, the company said, it suspended both models for all users, beginning at 5:21 p.m. Eastern time. Access to Anthropic’s other models was unaffected.
The order followed reports of a technique for bypassing Fable 5’s safety guardrails. Anthropic said its own review found the method surfaced only minor, already-known issues that other widely available models can also produce, and it disagreed with the basis for the order while complying with it.
In its reporting on the directive, Al Jazeera noted it was one of the few instances of the US government compelling a company to withdraw a deployed commercial AI model.
On June 30, Anthropic said the Commerce Department had lifted the controls and that both models were available again. The episode left open the question it raised: what standard of risk justifies a government forcing a deployed model offline, and how a narrow, contested jailbreak should weigh against a model already in use by millions.