Anthropic, the artificial intelligence company now valued at close to $1 trillion, received a export control directive from the Trump administration late Friday, ordering it to suspend access to its latest Claude models for all foreign nationals. The directive marks the second regulatory clash this year for the safety-focused firm and signals a sharp escalation in how Washington is willing to police advanced AI systems.
The administration cited "national security authorities" but did not specify a threat, Anthropic said in a blog post. The order applies to "any foreign national, whether inside or outside the United States, including foreign national Anthropic employees," effectively blocking the use of its Fable 5 and Mythos 5 models worldwide.
When safety advocacy provokes regulatory action
For years, Anthropic CEO Dario Amodei has publicly championed stricter government oversight of frontier AI. Just days before the directive, he published an essay arguing that "frontier AI models, like airplanes, should be required to go through technical testing and auditing, and their release should be blocked or reversed as a threat to public safety if they do not meet high standards of safety." The administration's move was not the kind of regulatory framework Amodei described.
Anthropic characterized the suspension as a "misunderstanding" and said it disagreed that the administration's finding warranted a recall. In its blog post, the company wrote, "As we have stated publicly, 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. This action does not adhere to those principles."
A voluntary order turns mandatory overnight
President Donald Trump's AI executive order, signed 10 days earlier, asked companies to voluntarily provide models for pre-release safety review. The new directive, sent to Anthropic on a Friday evening, carried the force of an export control mandate, leaving no room for negotiation. Daniel Remler, a senior fellow at the Center for a New American Security, told CNBC the difference in tone was stark. "This sure looks mandatory if there are going to be consequences for not doing what the government says," he said.
The Wall Street Journal reported that the suspension was prompted by conversations Amazon CEO Andy Jassy had with U.S. officials. Citing people familiar with the matter, the Journal said Jassy told the administration that Amazon researchers had used prompts to coax Fable 5 into producing information that could aid cyberattacks.
Anthropic had worked with government agencies to test the models before release and received approval to deploy them, according to a person familiar with the discussions. The person said there had been no prior communication from the government of a national security threat. The incident underscores the practical implications of the administration's use of export control authorities - an area of growing focus for many government AI policy teams.
Escalating tensions between AI labs and Washington
The directive arrives months after the Department of Defense designated Anthropic a supply chain risk, requiring defense contractors to certify they will not use Claude models in military work. Anthropic sued the Trump administration to reverse that designation; the litigation is ongoing. Defense Secretary Pete Hegseth wrote on X that the latest directive proves blacklisting Anthropic was "the right move."
David Sacks, the venture capitalist who previously served as Trump's AI and crypto czar, has long accused Anthropic of "running a sophisticated regulatory capture strategy based on fear-mongering." After the directive was announced, Sacks posted on X that the company's "minimizing language" was inconsistent with its brand as "the AI safety company," adding, "The ball is in Anthropic's court."
Alex Stamos, a former Facebook security chief now at Corridor, published an open letter addressed to Commerce Secretary Howard Lutnick and National Cyber Director Sean Cairncross, signed by more than 150 executives and technical leaders. The letter calls for the directive to be lifted and for rules that are "based upon science." Stamos told CNBC that the fear for every American AI company is that "at any moment you can now, if you run afoul of the administration, be shut down for a completely arbitrary decision capriciously."
Why this matters for government professionals
The sudden use of export control directives to block an AI model's deployment, outside any established rulemaking process, signals that Washington is willing to act swiftly when it perceives a national security risk. For government employees - whether in procurement, cybersecurity, national defense, or policy - this means that the AI tools available today could be restricted tomorrow without warning. It also places a premium on understanding how AI models are evaluated for safety and how interagency coordination on technology policy actually works.
The AI Learning Path for Policy Makers helps government professionals build a foundation in AI policy, threat modeling, and regulatory frameworks, providing the kind of technical and legal context these fast-moving disputes demand.
Your membership also unlocks:
