U.S. Court Rules Against Pentagon’s Ban on Anthropic’s Claude

Dynamic View Beating AI News: Anthropic has won this round of its legal battle with the Pentagon. U.S. Federal Judge Rita Lin ruled that the government’s designation of Anthropic as a “supply chain risk,” its directive to federal agencies to cease using Claude, and its restrictions on defense contractors cooperating with Anthropic were unlawful. The court found that these measures were retaliatory actions targeting Anthropic for its public criticism of government AI policies. The judge heavily criticized the government’s submitted evidence: the entire rationale for these actions was based on a single four-page memo, which was issued after two of the three measures had already been enacted. The government had previously expressed concerns that Anthropic could modify its models via “backdoors” after deployment, but trial materials confirmed Anthropic lacks such authority. The judge concluded that the government’s only remaining justification was mere “distrust” of Anthropic and a desire to make an example of it publicly. The conflict originally arose over restrictions on military use of Claude: Anthropic insisted its model could not be used for mass surveillance of Americans or fully autonomous weapons, while the Pentagon demanded unrestricted use for all lawful purposes. The court also clarified that the ruling does not compel the military to continue using Claude; the Pentagon may still legally switch vendors, but it can no longer ban Anthropic under the pretext of “supply chain risk.”