Anthropic AI models Fable 5 and Mythos 5 were taken offline after the U.S. government ordered the company to prevent access by foreign nationals, the company said Friday.
Anthropic said the directive did not specify the full national security concern. The company said it understood the order related to a possible jailbreak that could bypass safeguards on Fable 5.
The Associated Press reported that Anthropic took the models offline to comply with the Trump administration directive. AP said the move marked the U.S. government’s most significant step so far to restrict access to advanced AI models.
Anthropic launched Fable 5 and Mythos 5 earlier in June. The company said Fable 5 includes safeguards for areas such as cybersecurity, while Mythos 5 offers the same underlying model to vetted cyberdefenders and infrastructure providers.
Anthropic said Mythos 5 is available only to a small group of vetted partners due to its cybersecurity and biological capabilities. Its model page also said access to Claude Mythos 5 and Claude Fable 5 was unavailable for Glasswing partners.
The company said all other Anthropic models would remain available. It also said it was working to restore access to Fable 5 and Mythos 5 after what it called a misunderstanding. Anthropic asked it to revoke access to the models for users in all regions.
