Anthropic senior staff met with Trump administration officials in Washington, D.C., on Monday to try to resolve the artificial intelligence company’s latest high-profile dispute with the U.S. government, according to a person close to the company.
Anthropic received an export control order on Friday that names “national security authorities” and orders it to stop access to its latest AI models, Fable 5 and Mythos 5, “by foreign nationals, whether in the United States or abroad,” according to a statement.
The AI startup has disabled access to its models for all customers to ensure compliance with the directive.
The unexpected move marks the latest wrinkle in Anthropic’s relationship with the government since a clash with the Pentagon escalated earlier this year. In March, the Department of Defense designated Anthropic as a supply chain risk and barred defense contractors from using its technology, which it says threatens U.S. national security.
Secretary of Defense Pete Hegseth referenced the government’s latest directive in a post on Saturday’s X, writing that “every day that passes” proves why blacklisting Anthropic was the “right thing to do.”
Anthropic is suing the Trump administration to remove its supply chain risk designation, and the lawsuit is ongoing.
Anthropic announced Fable 5 and Mythos 5 on Tuesday, just days before receiving an export control directive from the government. The company worked with government agencies to test the model ahead of launch and receive approval for deployment, said a person familiar with the discussions, who requested anonymity to discuss confidential matters.
The government called Anthropic at 1 p.m. ET on Friday and told the company to disable Fable 5 and Mythos 5, citing unspecified national security threats, the person said. Anthropic received a formal letter asking for the model to be suspended at approximately 5:30 pm ET.
Before the directive was issued Friday, Anthropic had not received any communications regarding national security threats, the official said.
Fable 5 and Mythos 5 are built on Anthropic’s Claude Mythos Preview, a powerful model that excels at identifying security vulnerabilities in software. Anthropic has limited its rollout to a select group of companies as part of a cybersecurity initiative called “Project Glasswing,” and that approach appears to have received some goodwill from the Trump administration, which has held several talks with the company about the model’s capabilities.
Anthropic touted the Fable 5 and Mythos 5 as cutting-edge models that exceeded many industry benchmarks. While Mythos 5 is still limited to some users, Anthropic has made Fable 5 available to enterprise customers and paid subscribers. The company said the wide release was possible thanks to new safeguards that prevent responses in certain high-risk areas, such as cybersecurity and biology.
Anthropic said in a statement Friday that it believes the government’s concerns lie in the “limited and non-universal possibility of jailbreaking,” where users could circumvent cybersecurity guardrails and request Fable 5 to “read a specific codebase to fix flaws in the software.”
“We do not agree that the discovery of a narrow jailbreak possibility should be cause for a recall of a commercial model that has been deployed to hundreds of millions of people,” Anthropic said in a statement. “If this standard were applied industry-wide, we believe it would effectively halt the introduction of all new models for all Frontier model providers.”
Anthropic characterized the controversy as a “misunderstanding” and said it was working to restore access to its models “as soon as possible.”
Attention: Anthropic disables new Mythos class models days after release in response to government directives
