Anthropic has taken its newest and most powerful AI models, Mythos 5 and Claude Fable 5, offline worldwide following the Trump administration’s order imposing strict export controls on the systems for national security reasons. The directive required Anthropic to block access to the models to any foreign national, including those inside the United States and even the company’s own foreign staff. Anthropic says it cannot reliably separate foreign and non‑foreign users across its products, so it has shut down Fable 5 and Mythos 5 for everyone to comply.
Claude Fable 5 launched only days ago as Anthropic’s most powerful publicly accessible AI model. But some analysis found that a narrow jailbreak path still allows the system to analyze code for vulnerabilities, which critics said could aid offensive hacking in some scenarios. This led to the export control order.
The Trump administration had already ordered US federal agencies to phase out Anthropic’s tools from government work earlier this year after clashes over military use and safety constraints. Anyone who had routed tasks to Fable 5 now falls back to earlier Claude models.
Anthropic calls the situation a misunderstanding and says it hopes to restore access to Fable 5 and Mythos 5 as soon as it reaches an agreement with US authorities. In the meantime, no other
