Anthropic asked a federal appeals court in Washington Tuesday to reconsider its ruling that the Pentagon was entitled to brand the artificial intelligence company a national security supply-chain risk, Reuters reported. The request keeps alive a case that will help decide whether the government can shut a domestic technology vendor out of defense work because of the limits it places on how its products are used.

What Happened

Anthropic filed for rehearing at the U.S. Court of Appeals for the District of Columbia Circuit, which ruled against the company 2-1 in September, according to Reuters. Reuters did not say whether the company is asking the same three-judge panel or the full court to take another look.

In its filing, Anthropic argues that the panel majority misread the law the Pentagon relied on, turning a statute "aimed at hostile foreign actors into one that presumptively reaches every government contractor," Reuters reported. The company also says the court let the Pentagon act "without any pre-designation process," and that the label was retaliation for the company's public statements about AI safety.

RelatedMI5 Chief Says AI Is ‘Supercharging’ the Cyber Contest and Flags Bioweapon Risk

Anthropic has said the blacklisting has cost it billions of dollars in lost business. Reuters did not independently verify that figure. Reuters did not report a Pentagon response to the filing.

Why It Matters

A supply-chain risk designation is meant to keep products that officials consider a security threat out of defense systems. In Anthropic's case, the Associated Press reported, the label lets the Pentagon keep removing the company's Claude models from its systems and bars the use of Anthropic products in Defense Department work, including by contractors, though not across the whole federal government.

That reach to contractors is what makes the case significant beyond one company. Any business that sells into the defense supply chain and uses AI tools has to track which vendors are permitted. And any technology company that sets acceptable-use rules for government customers, such as bans on particular military applications, now faces a court ruling that such rules can be treated as a security risk. Boards of AI developers weighing government contracts will want to know whether usage policies can be enforced without risking exclusion.

Context

The dispute began in February, when President Donald Trump and Defense Secretary Pete Hegseth accused Anthropic of endangering national security and designated it a supply-chain risk, the AP reported. Anthropic had refused to let its AI be used for autonomous weapons or mass surveillance, according to Reuters.

In the Sept. 25 decision, Judge Gregory Katsas, joined by Judge Neomi Rao, wrote that the Pentagon had "ample support" for its actions and reasonably feared Anthropic could manipulate Claude to block national security functions, the AP reported. The judges said they had no reason to doubt Anthropic's intentions but that a bad motive was not required for a designation. Judge Karen LeCraft Henderson dissented. The majority also rejected Anthropic's retaliation claim, Reuters reported.

In a separate case in August, a federal judge in California struck down a parallel designation the Pentagon made under a different law, according to Reuters. The AP reported that the judge found the Pentagon acted illegally by punishing Anthropic for criticizing its views on AI use. The D.C. Circuit panel said it had no quarrel with that decision.

The litigation has run alongside other contact between the company and the administration, including chief executive Dario Amodei's one-on-one dinner with Trump in late September.

What to Watch

Reuters did not say when the court might act on the petition. Watch whether the D.C. Circuit asks the government to respond, and whether the case heads toward the Supreme Court if rehearing is denied. Any change in the Pentagon's position, or in how defense contractors are told to treat Anthropic products, would also shift the practical stakes.