Anthropic was already in a tense dispute with the federal government over its dealings with the Pentagon when it was hit with another major setback: a June 12 order requiring the company to cut off foreign access to its newly released AI models, Fable 5 and Mythos 5.
When the models debuted on June 9, Anthropic described Fable 5 as more capable than any system it had previously made broadly available. The company also said Claude Mythos 5 was built on the same core model, but with some safeguards loosened.
Reports say the government directive followed discussions between Amazon and the White House, after researchers claimed they had discovered methods for prompting Fable 5 to provide information that could potentially be used in cyberattacks.
In response, Anthropic suspended access to both models for all users. In a statement, the company said it was complying with the government’s legal order by removing Fable 5 and Mythos 5 from public availability. At the same time, Anthropic pushed back on the rationale, arguing that a limited potential jailbreak should not justify recalling a commercial model already deployed to hundreds of millions of people.
