A federal judge has ruled that the Pentagon's designation of AI company Anthropic as a "supply chain risk" was unlawful, delivering the company its first courtroom win in a monthslong fight with the Trump administration over the label.

U.S. District Judge Rita Lin of the Northern District of California found that Defense Secretary Pete Hegseth's decision to brand Anthropic a supply chain risk amounted to "unlawful retaliation" in violation of the First Amendment, was "arbitrary and capricious," and denied the company due process under the Fifth Amendment, according to NPR's report on the ruling.

How the fight started

The dispute traces back to February, when Hegseth took the unprecedented step of formally labeling Anthropic a supply chain risk — a designation that, according to TechCrunch's reporting, had previously been reserved almost exclusively for companies with ties to foreign adversaries. The label meant no part of the Defense Department, including its outside contractors, could do business with Anthropic's products.

The move came after Anthropic refused a Pentagon request to strip out internal safety guardrails on its Claude models — restrictions the company had built specifically to prevent military use of its AI for autonomous weapons systems or mass surveillance programs. Anthropic sued, arguing the designation was retaliatory rather than a genuine security determination.

THE DESIGNATION HAD PREVIOUSLY BEEN USED ONLY FOR COMPANIES CONNECTED TO FOREIGN ADVERSARIES

In her ruling, Judge Lin ordered the administration to rescind the designation entirely and to withdraw its directive telling federal agencies to cut ties with the company, according to Forbes' coverage of the decision. The ruling doesn't resolve the broader policy question of whether the Pentagon can require AI vendors to loosen safety restrictions as a condition of doing business with the military — only that this particular tool, the supply-chain-risk label, was used improperly here.

The Pentagon had not issued a public response to the ruling by press time, and it remains unclear whether the Justice Department will appeal. The case is being closely watched across the AI industry, where several companies have faced pressure from federal agencies over how much their models can be restricted from military and intelligence use cases.

For now, the practical effect is that federal agencies and contractors that had been barred from using Anthropic's Claude models under the designation are expected to be able to resume those relationships once the rescission takes effect, though the timeline for that process hasn't been spelled out in the ruling as reported.