The U.S. government directed the artificial intelligence company Anthropic to block all non-Americans from using two of its most powerful models, the firm announced Friday. Anthropic said in a statement that it would cut off all customers’ access to the models—Fable 5 and Mythos 5—to ensure compliance with the unprecedented directive.

Mythos has alarmed U.S. officials because of its ability to exploit cybersecurity vulnerabilities, helping drive the Trump administration to issue an executive order earlier this month requesting that AI firms grant the government early access to cutting-edge models. 

Subscribe Today

Get daily emails in your inbox

Email Address:

Commerce Secretary Howard Luttnick had sent a letter to Anthropic’s CEO Dario Amodei on Friday informing him that the models would be subject to export controls, Axios reports. A White House official told Axios that another company claimed it had succeeded in bypassing the Mythos AI’s safeguards, raising national security concerns.

Anthropic said the export control directive applies to “any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.”

President Donald Trump had previously ordered government agencies to stop using Anthropic’s models after the company rejected demands by the Defense Department to provide unrestricted access.