Judge Strikes Down Pentagon Anthropic Ban

A federal court ruled the Trump administration's Anthropic blacklist was unlawful First Amendment.

On Thursday, U.S. District Judge Rita Lin of the Northern District of California ruled that the Trump administration acted illegally when it designated Anthropic a national security supply chain risk and ordered federal agencies to stop using the maker of Claude. She wrote that the measures were "illegal and baseless" and that the "empty invocation of national security is not a blank check to punish and retaliate against government critics."

The designation was the first time a U.S. company had been publicly branded a supply chain risk under an obscure procurement statute designed to protect military systems from foreign sabotage — a label normally reserved for foreign firms. Anthropic was, until the feud, the only major AI company working with the Pentagon on classified systems and had held high-level security clearances.

The fight began in fall 2025, when the Pentagon asked Anthropic to bring Claude to its GenAI.mil platform but pressed for terms permitting "all lawful uses." Anthropic refused, holding two red lines: no use of Claude for mass surveillance of Americans or fully autonomous lethal weapons.

The government's justifications shifted repeatedly. According to Straight Arrow, an internal Pentagon memo claimed Anthropic could "disable" or alter its models mid-military-operation, but Under Secretary Emil Michael dropped that claim in a later sworn declaration, and the judge found no evidence Anthropic can access a model once deployed. Lin also found the actions violated the Fifth Amendment's due process guarantee, and wrote that officials publicly branded Anthropic a threat while privately emailing Amodei that a new deal was "very close" — evidence, she said, of retaliatory intent.

Anthropic warned the blacklist could cost it billions in lost revenue and reported fearful inquiries from more than 100 customers, with a two-year, up to $200 million Pentagon contract at risk, per Straight Arrow. The ruling is being watched across the tech industry: the Computer & Communications Industry Association said the case matters to anyone doing business with the U.S. government. Reuters notes Anthropic is also planning an IPO later this year that could value the firm at $2tn and raise $100bn.

The ruling does not force any agency to buy Anthropic's products, and the administration could appeal or pursue lawful procurement alternatives. A second Anthropic lawsuit remains pending before the D.C. Circuit Court of Appeals, where Reuters reports two of the three judges are Trump appointees skeptical of the company's arguments — and CNBC reports that panel must also rule for Anthropic before the Pentagon's Claude ban is officially lifted. Anthropic welcomed the ruling, saying it remains focused on working productively with the government on national security AI.