Anthropic said it suspended access to its Mythos 5 and Fable 5 models after a U.S. government directive restricted foreign-national access, while disputing the government’s stated jailbreak rationale. Axios, Reuters via The Guardian, and MIT Technology Review framed the move as part of a broader dispute between the...
The U.S. government ordered restrictions on foreign-national access to Anthropic’s Mythos 5 and Fable 5 models, prompting the company to disable both systems for all customers.
Anthropic said in a company statement that a U.S. export-control directive required it to suspend access to Fable 5 and Mythos 5 for foreign nationals. The company said it responded by disabling both models for all customers, rather than attempting to enforce access limits by nationality across its user base.
Axios reported that Commerce Secretary Howard Lutnick sent Anthropic a letter placing Mythos 5 and Fable 5 under export controls. Reuters, in a report carried by The Guardian, also reported that Anthropic disabled its most advanced AI models after a U.S. order limiting foreign-national access.
MIT Technology Review described the episode as the latest development in an ongoing feud between Anthropic and the U.S. government. The publication reported that Anthropic had announced an AI model called Mythos in April, and framed the latest dispute as one of several issues to watch in the company’s relationship with federal authorities.
In its statement, Anthropic said the directive was tied to concerns about jailbreaks, but the company disputed that rationale. A jailbreak generally refers to a technique for bypassing a model’s safety restrictions or intended limits. Anthropic said it disagreed with the basis for the action while still complying with the directive by suspending access.
The company’s public position matters because the dispute is not only about whether specific users can access particular models. It also touches on how the U.S. government may classify frontier AI systems under export-control rules, and how companies are expected to enforce those rules in practice.
Export controls are traditionally used to restrict access to technologies that the government considers sensitive for national security reasons. In this case, Axios reported that the Commerce Department placed Anthropic’s models under such controls, while Reuters via The Guardian tied the action to limits on foreign-national access.
That distinction is important. The reported order did not simply target a foreign country or a single overseas customer. According to the cited reports and Anthropic’s statement, it restricted access by foreign nationals, a category that can include people located inside the United States as well as abroad. Anthropic’s decision to disable the models for all customers suggests the company viewed selective enforcement as difficult or impractical under the directive.
Reuters via The Guardian reported that the model shutdown is connected to a wider dispute between Anthropic and the government over surveillance and weapons restrictions. Axios similarly described the letter from Lutnick as an escalation in Anthropic’s running fight with the Trump administration.
MIT Technology Review’s coverage presented the issue as a set of developments to watch rather than a settled policy outcome. Based on the available reports, the unresolved questions include whether the government will apply similar controls to other advanced AI systems, how companies will verify who is allowed to use restricted models, and whether Anthropic’s challenge to the jailbreak rationale will influence future decisions.
The immediate consequence is clear: Anthropic says Fable 5 and Mythos 5 are disabled for all customers after the U.S. directive. The longer-term consequences are less certain.
If the Commerce Department continues to treat advanced AI models as controlled technologies, other companies may face similar compliance questions. If Anthropic continues to dispute the reasoning behind the order, the conflict could also shape how model safety concerns are weighed against export-control authority.
For now, the public record shows a company complying with a government order while openly contesting its premise, and news organizations including Axios, Reuters via The Guardian, and MIT Technology Review treating the episode as a significant test of U.S. AI governance.
government ordered restrictions on foreign national access to Anthropic’s Mythos 5 and Fable 5 models, prompting the company to disable both systems for all customers.
What happened Anthropic said in a company statement that a U.S.
export control directive required it to suspend access to Fable 5 and Mythos 5 for foreign nationals.
Continue reading
