Federal Judge Rules Pentagon’s Blacklisting of Anthropic Was Illegal

A federal judge has ruled that the Pentagon’s campaign against Anthropic, which barred federal agencies and defense contractors from using the company’s Claude AI models, was illegal retaliation for Anthropic’s public criticism of the Department of Defense’s approach to military AI use.

U.S. District Judge Rita Lin, an appointee of former President Joe Biden who sits on the U.S. District Court for the Northern District of California, issued a 59-page written order on the night of Thursday, August 27, granting key portions of Anthropic’s motion for summary judgment. “The undisputed record shows that the challenged actions constituted unlawful retaliation in violation of the First Amendment,” Lin wrote, vacating the government’s directives and ordering the Trump administration to rescind them.

How the Dispute Started

The fight began in February, after Anthropic refused to remove safeguards that kept Claude from being used for fully autonomous weapons and mass domestic surveillance. At a meeting with the company, Defense Secretary Pete Hegseth called Claude’s capabilities “exquisite” and said Anthropic’s concerns were “understandable,” but gave the company three days to agree to “all lawful uses” of its technology or be designated a supply chain risk. Hegseth also raised the possibility of invoking the Defense Production Act, the Korean War-era law that lets the government force companies to prioritize contracts it deems essential to national security.

When Anthropic held its ground, Trump and Hegseth designated the company a supply chain risk, ordered every federal agency to permanently stop using its products, and banned defense contractors from doing any business with Anthropic at all, even work unrelated to the military. Anthropic CEO Dario Amodei refused to back down, citing concerns that the company’s tools could be used for mass surveillance or autonomous armed drones. Anthropic sued the Pentagon in March, calling the campaign an “unlawful campaign of retaliation.”

The case had already gone partway in Anthropic’s favor before Thursday’s ruling. Lin had temporarily blocked the supply chain risk designation and blocked enforcement of a directive Trump issued on social media ordering federal agencies to stop using Anthropic and Claude. At a hearing on July 30, she called the government’s position “really troubling.”

What the Judge Found

Lin’s final ruling went further than those earlier, temporary wins, finding that the government’s own record undercut its stated national security justification. According to The Register, which reviewed the full order, the judge found that the Pentagon’s national security rationale was assembled only after Hegseth had already decided Anthropic was a threat, and that some of the government’s central claims about what Claude could actually do were “entirely unfounded.”

Lin’s order refers throughout to the “Department of War,” which Ars Technica described as the Trump administration’s preferred name for the Department of Defense. “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. “The empty invocation of national security is not a blank check to punish and retaliate against government critics,” she added. According to the Associated Press, republished by SecurityWeek, Lin also wrote that the government’s actions “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.” Neither the Constitution nor the federal statute the government invoked, she added, allows officials to “impose sweeping penalties based principally on Anthropic’s critique of the Administration’s views.”

What Happens Next

An Anthropic spokesperson welcomed the ruling, calling the supply chain risk designation unlawful and saying, in a statement quoted by Ars Technica: “We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology.” The White House did not immediately respond to a request for comment, and the government is expected to fight the ruling.

The fight is not entirely over. Anthropic still has a separate, narrower case pending in the federal appeals court in Washington, D.C., challenging a different rule the Pentagon has used to try to declare the company a supply chain risk. And despite the ban, government agencies kept using Anthropic’s tools even as the two sides gradually mended ties ahead of the company’s planned initial public offering, Reuters reported, according to Ars Technica’s coverage.

The episode also underscores how differently Anthropic and its biggest rival have handled Washington. Hours after the government punished Anthropic for refusing to loosen its restrictions, OpenAI struck its own deal to work with the Pentagon. Both companies are now preparing for high-profile IPOs, giving each an added incentive to keep its government relationships intact even as they diverge on how much to accommodate military use of their technology.