Microsoft and two sets of outside supporters moved to deepen their fight in federal court over the Pentagon’s action targeting Anthropic, with the software giant and retired senior military officers urging a judge to pause what they describe as a damaging “supply chain risk” designation.
The latest filings come in a case where Anthropic sued the Trump administration on Monday, arguing that the designation and related Pentagon steps should be blocked. Microsoft said in its own Tuesday filing that it is challenging Defense Secretary Pete Hegseth’s action last week after the Pentagon labeled Anthropic’s AI products as posing a threat to national security and used that determination to effectively shut Anthropic out of military work.
Microsoft, which is a major government contractor, argued that using a supply-chain-risk designation to resolve a contract dispute risks significant harm beyond the public interest. The company asked the judge to order a temporary lifting of the designation to allow for “reasoned discussion” between Anthropic and the Trump administration.
In parallel, a group of 22 retired high-ranking military officials backed Anthropic’s position, including former CIA director Michael Hayden, who retired as an Air Force general, and retired Coast Guard Adm. Thad Allen. In their court filing, the officials said Hegseth’s conduct reflects a misuse of government authority for “retribution against a private company that has displeased the leadership.”
The dispute has played out publicly amid a broader conflict between the Pentagon and Anthropic over military access to the company’s Claude model. The Pentagon took action against Anthropic after what the reporting described as an unusually public disagreement over whether Anthropic would allow unrestricted military use of its AI model.
Trump also said he was ordering all federal agencies to stop using Claude, escalating the pressure and contributing to the current court challenges. While the Pentagon declined to comment on the litigation, the filings from supporters laid out arguments that a government determination framed around national security could be used as leverage in what supporters portray as contract negotiations.
Microsoft’s filing also said it supports Anthropic’s “two ethical red lines” that were a sticking point during negotiations after the Pentagon insisted the company must allow for “all lawful” uses of its AI. Microsoft said it believes “American AI should not be used to conduct domestic mass surveillance or start a war without human control,” adding that it views that position as consistent with law and broadly supported by American society.
The federal case is being overseen by U.S. District Judge Rita Lin, who was nominated to the bench by President Joe Biden in 2022. Lin has scheduled a March 24 hearing in the San Francisco matter, where Anthropic is headquartered, and Anthropic has also filed a separate, narrower suit in the federal appeals court in Washington, D.C.
Support filings in the case included additional submissions from the AI industry, including a filing from a group of AI developers at Google and OpenAI, and another from organizations such as the Cato Institute and the Electronic Frontier Foundation. The retired military officials’ filing warned that the “sudden uncertainty” created by targeting a technology embedded in military platforms could disrupt planning and potentially put soldiers at risk during ongoing operations.
Separately, the current commander of U.S. Central Command confirmed in a video posted to social media Wednesday about U.S. strikes on Iran that military forces were using “advanced AI tools” to “sift through vast amounts of data in seconds,” while also saying humans “will always make final decisions” on what to shoot and when. The reported reference did not name which tools were being used, and the lawsuit-related filings did not mention Iran, according to the account.
As the legal fight proceeds, military officials said they were looking to shift classified-network work away from Anthropic to competitors including Google, OpenAI and Elon Musk’s xAI, after Anthropic was singled out amid the dispute over access and permitted use.