Anthropic announced on Friday that it has suspended access to its advanced AI models, Fable 5 and Mythos 5, following a US government directive citing national security concerns.
The decision came just three days after the launch of Fable 5. In a blog post, the company revealed that it had received an order prohibiting all foreign nationals – including Anthropic employees – from accessing either model.
“As a result of this directive, we are required to immediately disable Fable 5 and Mythos 5 for all customers to remain compliant,” the company stated.
According to Anthropic, the notification was received at 5:21 p.m. local time (2121 GMT). Axios reported that the directive was issued by US Commerce Secretary Howard Lutnick.
US Commerce Department officials did not immediately comment on the matter.
Anthropic said the order did not specify the exact reason behind the restriction. However, the company believes the government may have identified a technique for bypassing, or “jailbreaking,” Fable 5’s safeguards, potentially enabling its use in cyberattacks.
Launched on Tuesday, Fable 5 is a restricted version of Mythos 5, a highly advanced AI system that Anthropic had previously withheld from public release due to concerns over its exceptional ability to detect software vulnerabilities that could be exploited by hackers.
Mythos 5, the unrestricted version of the AI model, has been made available only to a limited number of companies.
Anthropic stated that it reviewed the alleged “jailbreaking” technique and the potential security risks associated with it. However, the company maintains that Fable 5 does not provide hackers with capabilities beyond those already accessible through other publicly available AI models.
According to Anthropic, its security teams have not identified any “universal jailbreak” capable of consistently bypassing the model’s safeguards designed to prevent malicious cyber activity.
“We do not believe that the discovery of a limited potential jailbreak warrants the withdrawal of a commercial model used by hundreds of millions of people,” the company said.
Anthropic further argued that applying such a standard across the AI industry would significantly hinder innovation and could effectively prevent frontier AI developers from releasing new models.
The company has also been engaged in an ongoing dispute with the Trump administration after refusing to permit its technology to be used for mass surveillance programs or autonomous weapons systems. The disagreement reportedly resulted in the Pentagon ending its contracts with Anthropic.