Federal Judge Throws Out Pentagon's Retaliation Campaign Against Anthropic
A federal judge has handed Anthropic a significant legal win, ruling that the Pentagon's decision to designate the AI company a supply chain risk was unlawful and without legitimate foundation.
U.S. District Judge Rita Lin issued her 59-page ruling last Thursday, finding that the Department of Defense acted illegally when it punished Anthropic for publicly criticising the government's approach to AI in military applications. The government is expected to appeal.
The whole mess started in February when President Trump and Defence Secretary Pete Hegseth accused Anthropic of being a national security liability. The trigger was Anthropic CEO Dario Amodei's refusal to drop the company's objections to its technology being used for mass surveillance or autonomous weapons systems. Rather than engaging with those concerns, the administration opted to make an example of him.
Lin was blunt about that in her ruling. The government's actions, she wrote, "were based on a desire to make a public example out of Anthropic for its 'arrogance' in criticizing the government, not based on any articulable basis to believe that Anthropic would actually sabotage its model." That is a fairly damning summary of how the world's most powerful military chose to handle a corporate policy disagreement.
Neither the Constitution nor the federal statute the Pentagon invoked permits the government to "impose sweeping penalties based principally on Anthropic's critique of the Administration's views," Lin added. Hard to argue with that, though the DOJ will presumably try.
For its part, Anthropic issued a statement welcoming the ruling and promising to work productively with government on national security applications of AI. Diplomatically worded, given the circumstances.
Anthropologic filed suit against the Pentagon in March, describing the supply chain designation as an "unlawful campaign of retaliation." Earlier in proceedings, Lin had already temporarily blocked both the supply chain label and a Trump directive ordering all federal agencies to stop using Anthropic's Claude chatbot. At a July hearing she said the government's position was "really troubling" and appeared "at odds with the First Amendment" - and that the record had only gotten worse for the government as the case progressed.
DOJ lawyers tried a novel argument during that hearing, claiming AI models are "so staggeringly enormous and opaque" that the Defence Department cannot assess them the way it would physical hardware. Whether that logic resurfaces on appeal remains to be seen.
The timing of all this is worth noting. Just hours after the Pentagon moved against Anthropic, OpenAI quietly announced its own deal to work with the DoD. Both companies are heading toward IPOs. The contrast in their respective strategies toward Washington could not be more visible.
A separate, narrower case brought by Anthropic involving a different Pentagon rule is still working its way through the federal appeals court in Washington. So this particular fight is not entirely over yet.