A US appeals court denied Anthropic’s request to pause the Pentagon’s 'supply chain risk' designation, dealing a blow to the AI firm. While a California court earlier blocked similar action, the latest ruling allows the Pentagon to restrict new contracts. The dispute stems from Anthropic refusing AI use in weapons and surveillance.

A federal appeals court in Washington, DC denied AI company Anthropic's request to pause the Pentagon's designation of the firm as a supply chain risk, delivering a blow to the Claude developer in its ongoing legal battle with the Trump administration. The Wednesday ruling contrasts sharply with an earlier California court decision that sided with Anthropic.

The DC court's decision, reported first by Reuters, comes after a California federal judge blocked a separate Pentagon order in late March, saying the department appeared to have unlawfully retaliated against Anthropic for its AI safety views. The split decisions mean Anthropic still faces significant legal hurdles despite its earlier courtroom victory

The San Francisco injunction means non-Pentagon agencies no longer have to terminate contracts with Anthropic, but the DC decision means the Pentagon can still treat the company as a supply-chain risk and exclude it from new Pentagon contracts.

AI safety dispute