Federal judge rules Pentagon's supply-chain risk label against Anthropic illegal

What the court decided

A federal judge in California has ruled that the U.S. government's designation of Anthropic as a supply-chain risk was illegal, handing the AI company its first court victory in the dispute. According to TechCrunch, U.S. District Judge Rita Lin issued the ruling on Thursday evening, finding that Defense Secretary Pete Hegseth's classification of Anthropic as a national security threat constituted "unlawful retaliation" in violation of the First Amendment. She also described the decision as "arbitrary and capricious" and concluded that Anthropic had been denied due process under the Fifth Amendment.

How the dispute began

TechCrunch reports that earlier this year Hegseth and President Donald Trump labeled Anthropic a supply-chain risk and ordered every federal agency, including those with no defense mandate, to stop working with the maker of the Claude models.

The confrontation traces back to safety conditions Anthropic attached to its products. The company drew firm lines around the use of its models for fully autonomous weapons and for mass surveillance of American citizens. The Pentagon maintained that it intended to use the models only for lawful purposes and accused Anthropic of trying to dictate how the military deployed software it had already bought and paid for.

The judge's reasoning

Lin found that the government's own conduct undercut its national security justification. As TechCrunch reports, she wrote that the government's "words and deeds confirm that the challenged actions were based on a desire to make a public example out of Anthropic for its 'arrogance' in criticizing the government."

She pointed to several contradictions in the administration's position. Hegseth had proposed applying the Defense Production Act to Anthropic, a step that would treat the company as essential to national security rather than a threat to it. At the same time, the Department of Defense continued pursuing a contract with Anthropic, and the government was collaborating with the company's new model, Mythos, on cybersecurity. Lin also noted that Anthropic "undisputedly lacks" any backdoor access to its technology once it is handed over to the department.

"Though the Department of War is undisputedly free to select the AI vendor of its choice, the evidence demonstrates that the broad measures imposed on Anthropic were illegal and baseless," Lin wrote, adding that "the empty invocation of national security is not a blank check to punish and retaliate against government critics."

What happens next

Anthropic welcomed the outcome. In a statement shared with TechCrunch, a company spokesperson said Anthropic remains focused on working productively with the government to harness AI for national security so that all Americans benefit from the technology.

The ruling resolves only one front of the legal fight. Anthropic filed two complaints against the Department of Defense in March, one in California and one in Washington, D.C. The California case produced this week's decision, while the D.C. case is still pending. TechCrunch says it has contacted the DoD for comment on the ruling.

Why it matters

The decision is a significant check on how the U.S. government can wield procurement and supply-chain mechanisms against AI vendors. If it stands, it signals that federal agencies cannot simply invoke national security to sideline a contractor that imposes usage restrictions or criticizes policy, and that courts will examine whether such labels are consistent with the government's actual behavior toward the company.

The case also exposes a broader tension in public-sector AI adoption. Vendors increasingly attach safety conditions to frontier models, while government buyers expect full control over software they purchase. Judge Lin's ruling suggests the government remains free to choose whichever vendor it prefers, but not to punish one for where it draws its ethical lines. With the D.C. complaint unresolved and the possibility of an appeal ahead, the final boundaries of that relationship have yet to be settled.