The Trump administration pressured Anthropic to pull its latest artificial intelligence models offline, and the company said Friday it complied with a U.S. Commerce Department directive that barred non-Americans, including Anthropic’s employees, from accessing them. The move, paired with Defense Secretary Pete Hegseth’s attempt to designate Anthropic a supply chain risk after a Pentagon contract dispute, has raised alarms that the federal apparatus is now reaching directly into deployed AI systems and deciding who gets access, who gets shut out, and which tools remain usable.
Who Gets to Decide
Anthropic said it took its latest models, Fable 5 and Mythos 5, offline to comply with the directive. The company said it did not believe the government’s steps were warranted by the security concern it had flagged. Anthropic also said it had limited use of some of its latest technology to select customers because of its ability to surpass human cybersecurity experts in finding and exploiting computer vulnerabilities. The San Francisco-based company had previously discussed the latest models’ capabilities with the White House.
The Commerce Department did not immediately respond to a request for comment on Monday. That silence sits alongside a directive that, according to TechCrunch, effectively forced Anthropic to pull its latest AI models offline just before the weekend and showed that the AI industry is not immune to government interference. TechCrunch said the government’s action appeared to require no court approval.
What the Powerful Call “Risk”
Friday’s directive came 10 days after President Donald Trump signed an executive order to establish a framework for the federal government to vet the national security risks of the most advanced AI systems for up to a month before their public release. Participation by AI developers would be voluntary, the order said. In practice, the latest move shows how quickly voluntary language can sit beside hard-edged enforcement when the state decides to flex.
Tensions have been running high between the Trump administration and Anthropic, which has sought to put guardrails on the development of AI to minimize potential risks and maximize its economic and national security benefits for the U.S. After a contract dispute with the Pentagon, Defense Secretary Pete Hegseth sought to declare Anthropic a supply chain risk, an unprecedented move against a U.S. company that Anthropic has challenged in two federal courts. The company said it wanted assurance the Pentagon would not use its technology in fully autonomous weapons and the surveillance of Americans. Hegseth said the company must allow for any uses the Pentagon deemed lawful.
That is the hierarchy in plain view: a company asks for limits on autonomous weapons and surveillance, and the Pentagon answers with demands of its own, backed by the machinery of federal power and litigation in two federal courts.
Who Pays for the Crackdown
More than 100 cybersecurity experts and leaders from companies including Adobe and Nvidia asked the U.S. government in a letter Sunday to lift the export control directives on the Anthropic models and to “commit to an open, scientific and transparent process of handling AI risk assessments in the future.” The letter said Anthropic’s Mythos models are “quite good” at finding flaws in software and weaponizing exploits, but are “not uniquely good at these tasks,” and said many signatories regularly use other foundation and open-source models for security audits and training.
The letter said it was dangerous to take away the best cyber defense capabilities “without a good reason” when America’s adversaries are rapidly advancing. It said China’s models are “only months behind the best American models,” and that it is even likely China’s government has access to private capabilities beyond what has been made public. Katie Moussouris, a cybersecurity veteran and researcher who founded Luta Security, said in a blog post that the bypass described by security researchers “should never have triggered an export control.” Moussouris said, “The behavior described in the paper cannot meaningfully be fixed, and any attempt would only weaken the model for defense.”
TechCrunch said Moussouris and dozens of other top security researchers and experts called on the Trump administration to revoke the export control order, calling the move to pull advanced cybersecurity capabilities from network defenders in the U.S. “dangerous.” Justin Hendrix, the editor of Tech Policy Press, said the Trump administration’s move is “likely to raise alarms in foreign capitals about the reliability of American AI for critical applications.” He said the message is that AI companies in the United States cannot be trusted to operate without interference from the U.S. government.
The Washington Post said pressuring Anthropic to retract its Fable model had established a potential precedent for intervening in deployed AI models through export control authorities. It said Amazon CEO Andy Jassy first raised concerns about a Fable jailbreak with the White House and that the development suggested Amazon’s strategic partnership with Anthropic was becoming more complex. The Post also quoted Joe Khawam of the Law Reform Institute as saying the administration is likely to enforce export controls on AI models based on what they can say rather than where they are used.
Anthropic released Fable widely last week. Fable is a limited version of the more advanced Mythos, to which the company has tightly limited access because of cybersecurity fears. Anthropic said it took Fable 5 and Mythos 5 offline on Friday, June 12, 2026, to comply with the directive. The company’s latest clash with the federal government now stands as a warning shot: the state can reach into deployed systems, cut access, and call it security while everyone else scrambles to live with the consequences.