Anthropic suspends AI models after U.S. security directive

Anthropic abruptly disabled access to two of its newest artificial intelligence models Friday after the U.S. government ordered the company to prevent any foreign national from using them, citing national security authorities.

The company said the directive forced it to suspend Fable 5 and Mythos 5 for all customers, not only foreign nationals, because it could not otherwise ensure compliance. Fable 5, part of Anthropic’s Claude lineup, had been released to the public just days earlier and was described by the company as its most powerful generally available model.

“The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees,” Anthropic wrote in a statement posted late Friday.

“The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance,” the company said. Claude’s landing page said Friday evening that “Fable 5 is temporarily unavailable.”

NBC News reported that the directive was sent in a letter from Commerce Secretary Howard Lutnick and written with help from officials at the Commerce Department’s Bureau of Industry and Security, citing an administration official. The Commerce Department did not immediately respond to requests for comment from NBC News and the BBC.

Anthropic said U.S. officials appeared to be concerned about a technique that could bypass, or “jailbreak,” some safeguards in Fable 5. The company said the government had provided only “verbal evidence” and had not identified broader specific security concerns.

“Our understanding is that the government believes it has become aware of a method of bypassing, or ‘jailbreaking’ Fable 5,” Anthropic said, according to the BBC. The company said it reviewed a demonstration in which the technique was used to identify “a small number of previously known, minor vulnerabilities,” adding that the issues appeared “relatively simple” and that other publicly available models could find them without a bypass.

“We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people,” Anthropic said.

Anthropic released Fable 5 and Mythos 5 on Tuesday, saying the systems were powerful enough to require strict safety limits. “Releasing a model this capable comes with risks. Without safeguards, Fable 5’s capabilities in areas like cybersecurity could be misused to cause serious damage,” the company said in its launch announcement.

The two models were built on the same technical foundation, NBC News reported. Fable 5 was made available to the general public with stronger restrictions on questions involving areas such as cybersecurity and biology, while Mythos 5 was released without those safeguards to a select group of trusted partners, including cybersecurity and infrastructure companies.

The BBC reported that Anthropic had enabled pre-release access for a small number of organizations because the tool’s ability to exploit or hack computer systems made it potentially dangerous. The company had previously described the technology as “too powerful to release,” a phrase the BBC said some critics questioned as inflated hype and marketing spin.

The suspension comes amid an ongoing dispute between Anthropic and the Trump administration. In February, President Donald Trump and Defense Secretary Pete Hegseth moved to bar Anthropic’s products from federal agencies after the company sought stronger guardrails on how the Pentagon used its technology. Trump wrote on Truth Social that “The Leftwing nut jobs at Anthropic have made a DISASTROUS MISTAKE trying to STRONG-ARM the Department of War.”

The BBC reported that Hegseth labeled Anthropic a “supply chain risk,” a designation the outlet said historically has been reserved for companies based in adversarial countries. Anthropic sued the Pentagon over the designation. A U.S. judge has ruled the Pentagon directive cannot be enforced while the case continues, allowing government agencies and organizations working with the U.S. military to keep using Anthropic tools.

Anthropic said it believed Friday’s directive stemmed from “a misunderstanding” and that it hoped to restore access as soon as possible. The company said it supports a government process for blocking unsafe deployments, but objected to the way the order was issued.

“We believe the government should have the ability to block unsafe deployments, as part of a statutory process that is transparent, fair, clear, and grounded in technical facts,” Anthropic wrote. “This action does not adhere to those principles.”