A federal judge has blocked the Trump administration’s attempt to label Anthropic a national security risk, sharply limiting the government’s response to a dispute over how the company’s AI models could be used by the military.
The ruling, issued Thursday by US district judge Rita Lin in California, vacated the February 27 decision by defense secretary Pete Hegseth to designate Anthropic a “supply-chain risk.” That label had made the AI lab ineligible for federal contracts and restricted military contractors and suppliers from doing business with the company.
What The Court Decided
Lin found that the government’s actions amounted to unconstitutional retaliation. In the 59-page ruling, she wrote that the measures imposed on Anthropic were “illegal and baseless,” while also recognizing that the Department of War can choose which AI vendor it wants to use.
The decision removed the “supply-chain risk” designation and lifted an additional measure that had blocked contractors or suppliers to the US military from working with Anthropic. Lin described that added restriction as “arbitrary, capricious, an abuse of discretion, and otherwise not in accordance with law.”
The ruling also addressed penalties imposed by nine agencies, including the Pentagon, the Treasury Department, the State Department, and the Department of Homeland Security. Lin ruled that those agencies had improperly imposed sanctions on Anthropic, and the decision removes those penalties as well.
What The Ruling Does Not Do
The decision is not an order forcing the Pentagon to adopt Anthropic’s models. Lin affirmed that the Pentagon is not required to use Anthropic’s technology and may decide to use other models instead.
That distinction matters because the case was not simply about whether the government can choose one AI system over another. The court focused on whether the government could use a national security designation to broadly punish Anthropic after a disagreement over deployment limits.
A Pentagon spokesperson could not immediately be reached for comment, and the department is expected to appeal.
How The Anthropic Dispute Began
The case grew out of a conflict earlier this year between the Pentagon and Anthropic over a $200 million deal to use the AI lab’s Claude models for military applications.
The dispute intensified after reports that the US had used Claude in the operation to capture Venezuelan president Nicolás Maduro. After that operation, a Palantir employee relayed concerns from an Anthropic staffer to US officials about how the models had been used.
During negotiations, Anthropic sought limits on how its models could be deployed. Those limits included support for lethal autonomous weapons and mass surveillance systems.
Hegseth rejected those curbs. He argued that a contractor could not dictate how technology would be used once deployed and said the contract allowed “all lawful use.”
After negotiations broke down in February, Hegseth designated Anthropic a “supply-chain risk.” The Pentagon said at the time that the label was appropriate because giving Anthropic access to any classified systems would “introduce unacceptable risk” if the AI lab could disable or alter its technology, for instance, at a time of war.
Why The Judge Found The Label Unconvincing
Lin pointed to the government’s continued engagement with Anthropic as part of the reasoning behind the ruling. Anthropic’s top AI models are considered some of the best in the world, and its models recently became subject to the Trump administration’s AI oversight framework because of their powerful capabilities.
The judge cited the administration’s involvement in reviewing advanced Anthropic models. She wrote that the government was still discussing collaboration with Anthropic on its new model, Mythos, “in an array of sensitive contexts.”
Lin added: “None of that is consistent with a genuine fear that Anthropic is a saboteur who would poison its software to harm national security.”
For the court, that apparent contradiction weakened the government’s position. If officials were still considering sensitive collaboration with Anthropic, the broad blacklist looked less like a narrowly justified security measure and more like a response to the company’s stance in negotiations.
The Legal Fight Continues
Anthropic filed two lawsuits in response to the Pentagon’s actions. One was filed in federal district court in California, and the other was filed at the US Court of Appeals for the District of Columbia.
The company accused the Pentagon of violating its First and Fifth Amendment protections on ideological grounds. The DC case is ongoing.
In a statement, Anthropic spokesperson Danielle Cohen said, “We welcome the court’s ruling that this supply-chain risk designation was unlawful. We remain focused on working productively with the government to harness AI for our national security.”
The immediate effect is clear: Anthropic is no longer under the penalties removed by Lin’s ruling. The broader fight over government use of powerful AI models, vendor restrictions, and limits on military applications is not finished.