Appeals Court Upholds Pentagon Blacklist of Anthropic in 2-1 Ruling, Splitting With San Francisco Judge

The U.S. Court of Appeals for the District of Columbia Circuit on Sept. 25 upheld the Department of Defense's blacklisting of Anthropic, ruling 2-1 that the administration could designate the AI lab a supply-chain risk for withholding certain Claude features from the military . The panel found the Secretary "did not transgress any limits on his authority under the Supply Chain Security Act or the Constitution" and denied Anthropic's petitions for review . Judges Gregory Katsas and Neomi Rao formed the majority and rejected Anthropic's First Amendment retaliation claim, reasoning the company was excluded because it refused to agree to essential contract terms.

The dispute began when Anthropic declined to remove safety guardrails restricting autonomous lethal weapons and mass domestic surveillance uses from a $200 million military prototype contract. The same court denied Anthropic's emergency motion for a stay in April . The appellate decision now conflicts with an Aug. 27 ruling by U.S. District Judge Rita F. Lin in San Francisco, whose 59-page opinion found the designation to be unconstitutional First Amendment retaliation and blocked it.

Anthropic's founders, meanwhile, are asking shareholders to approve special shares giving CEO Dario Amodei and six co-founders a combined 50.1% of the vote on most matters, as long as at least three keep a minimum stake . The Motley Fool, citing reports, said the IPO has been pushed to November; Anthropic has not confirmed that timing . Partners are also in focus: AKAM announced a multibillion-dollar infrastructure deal with Anthropic, and Salesforce (CRM) CEO Marc Benioff has signaled he would likely sell the company's Anthropic stake, which is framed as a way to retire debt from its $25 billion share repurchase.

Watch Anthropic's next legal step, how the conflicting rulings are reconciled, and whether the designation appears in any IPO filing.