US restricts Claude Fable 5 access
On June 12, 2026, the United States government issued a national security export control directive ordering Anthropic to disable all access to its new Mythos-class AI models, Fable 5 and Mythos 5, for all foreign nationals both inside and outside the US, including Anthropic's own employees. In response to the order, which cited security concerns over a potential safety jailbreak method, Anthropic complied by disabling access to both models for all customers globally to ensure full compliance.
This unprecedented government action demonstrates the state's readiness to intervene directly and shut down access to frontier models under the guise of export controls and national security. While Anthropic has complied, they noted that the jailbreak method in question is minor and similar to vulnerabilities found in other public models.
* The suspension marks the first time a major LLM has been completely pulled from the market due to a government export control order.
* The restriction affects not just end users, but also Anthropic's own foreign national employees, presenting major internal operational hurdles.
* The incident underscores the fragility of commercial API dependencies for developers, as access to state-of-the-art models can be instantly revoked.
DISCOVERED
2h ago
2026-06-13
PUBLISHED
2h ago
2026-06-13
RELEVANCE
AUTHOR
aidenybai