A US appeals court has upheld the Trump administration's blacklisting of Anthropic technology in a 2-1 decision, ruling that the government had authority to blacklist the company for withholding certain AI features regardless of whether Anthropic acted with malicious intent. The ruling addresses a fundamental tension in military AI deployment: the government's concern that overly constrained models could fail during critical operations, weighed against Anthropic's concern that unconstrained systems might hallucinate inappropriate targets for lethal military force.

The decision was issued by the US Court of Appeals for the District of Columbia Circuit, with both judges in the majority appointed by Trump during his first administration. Notably, a California federal judge had previously ruled the blacklisting unlawful, but the appeals court reviewed the action under different legal authority—a more permissive procurement statute—rather than the stricter supply-chain security law the lower court examined. The blacklisting followed Trump and Defence Secretary Pete Hegseth's March directive to cease federal use of Anthropic's products, which prompted Anthropic's lawsuit. The company may pursue further appeals, potentially to the Supreme Court, though Commerce Secretary Howard Lutnick recently suggested the Trump administration and Anthropic have reconciled their differences.