Two hard limits were what Anthropic refused to give up in its dealings with the Pentagon: it would not enable mass surveillance of Americans, and it would not build lethal autonomous weapons. The consequence was a supply chain risk designation, a label typically reserved for actual national security threats. On Thursday, a federal judge declared that label illegal.
Judge Rita F. Lin of the northern district of California found the Pentagon’s move to blacklist the AI lab earlier this year unconstitutional. The decision hands a win for Anthropic in a fight with the Trump administration that has swung back and forth for months.
What the judge said
There was no hedging in Lin’s language. “The empty invocation of national security is not a blank check to punish and retaliate against government critics,” she wrote in the ruling.
The conduct amounted to “unlawful retaliation in violation of the First Amendment,” she said. Turning to Defense Secretary Pete Hegseth’s choice to brand Anthropic a supply chain risk, she wrote that it “was arbitrary and capricious. 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.”
That second sentence is worth a second look. Nowhere is the court dictating which AI supplier the Pentagon must use. What it forbids is punishing a vendor for speaking to reporters.
How a contract renegotiation turned into a lawsuit
The origins trace back to last winter, when Hegseth set out to renegotiate the existing military contract held by every AI lab so that the Pentagon could deploy AI for “any lawful use” — wording that quietly grants the department far broader authority than the previous terms allowed.
Signatures came in from most of the labs. Anthropic was the holdout, sticking to its two conditions: no mass surveillance of Americans, and no lethal autonomous weapons, by which it means AI systems capable of killing targets without human oversight.
Pressure then arrived from both sides — a stream of insults out of Department of Defense officials, capped by an ultimatum from the administration.
Amodei’s statement, and what came 24 hours later
Fewer than 24 hours before that ultimatum arrived, Anthropic CEO Dario Amodei made clear the company would not budge. Anthropic has “never raised objections to particular military operations nor attempted to limit use of our technology in an ad hoc manner,” he wrote, adding that in a “narrow set of cases, we believe AI can undermine, rather than defend, democratic values.”
The supply chain risk designation followed. The Pentagon simultaneously worked to dilute Anthropic’s standing inside the Department of Defense by inking agreements with seven rival AI labs, including Google, Microsoft, OpenAI and SpaceX.
The March order already told you where this was headed
March brought Anthropic’s lawsuit in a California district court, which alleged the administration had unlawfully retaliated against the company for drawing “red lines” — its phrase for the military applications of its technology that it will not accept.
Lin backed the company at that stage as well, putting a temporary block on the blacklist. “The Department of War’s records show that it designated Anthropic as a supply chain risk because of its ‘hostile manner through the press,'” she wrote in that order. “Punishing Anthropic for bringing public scrutiny to the government’s contracting position is classic illegal First Amendment retaliation.”
The government’s own documents, put differently, said the quiet part out loud.
Anthropic’s response is notably mild
A victory lap would be the obvious move. It never came.
“We welcome the court’s ruling that this supply chain risk designation was unlawful,” Anthropic spokesperson Danielle Ghiglieri said in a statement Thursday. “We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology.”
This is a company that prevailed in court and still wants the contract.