Anthropic cuts off access to Fable 5 and Mythos 5 after U.S. directive

Anthropic has disabled access to two of its flagship AI models, Fable 5 and Mythos 5, after receiving a directive from the U.S. government. According to the reports, the company said the action was taken to comply with the order, which was tied to export controls and cited national security concerns.

The move was abrupt and affected access to models that had been positioned as core parts of Anthropic’s AI offering. Both Cointelegraph and CNBC reported that the company suspended access rather than continuing to make the systems available under the same conditions. The decision underscores how quickly AI providers can be pulled into government oversight when policy and security questions intersect with advanced model deployment.

While the reports do not spell out the full contents of the directive, they indicate that the U.S. government framed the issue in terms of national security. That places the case within a broader pattern of heightened official scrutiny around frontier technologies, especially where access, distribution, and international controls may be involved. In this instance, the reported export-control basis suggests the government saw sufficient concern to require a change in access to the models.

Anthropic’s response, as described in the reports, was to comply with the directive by disabling access. The company did not appear to challenge the instruction in the source material provided, and no further details were included about whether the suspension was temporary or whether it would affect broader parts of its product lineup. The scope of the restriction, based on the available reporting, was limited to Fable 5 and Mythos 5.

The development is notable because it shows how regulatory and policy decisions can directly affect the availability of leading AI systems. For users and enterprises that rely on such models, sudden changes in access can create uncertainty about continuity, service planning, and compliance. For Anthropic, the action reflects the practical reality that companies operating at the edge of advanced AI development may need to adjust offerings in response to government directives.

The reports do not provide additional operational details, financial effects, or a timeline for any possible reinstatement of access. Even so, the episode highlights the growing intersection between artificial intelligence, national security, and export policy. As governments continue to examine the risks and controls around powerful AI systems, similar interventions may remain a feature of the sector’s operating environment.