US appeals court upholds Pentagon blacklist of Anthropic
A US appeals court has upheld the Department of Defense’s decision to blacklist Anthropic technology after the company withheld certain AI features from military use. In a 2-1 ruling, a panel of the US Court of Appeals for the District of Columbia Circuit found that the Trump administration had the authority to take that action even though Anthropic had no malicious intent.
The judges described the dispute as involving difficult questions about military applications of powerful AI systems. The ruling cited the government’s concern that models with excessive safeguards could shut down unexpectedly and disrupt important military operations. Anthropic, in turn, raised the risk that unrestricted systems could generate hallucinated or inappropriate targets for lethal force. The court said President Trump and Defense Secretary Pete Hegseth must weigh those competing risks, and concluded that the secretary had not exceeded his authority under the Supply Chain Security Act or the Constitution. The panel therefore denied the petitions for review. The same court had previously rejected Anthropic’s emergency request for a stay in April.
Written by Sakhanda Wire AI from reporting by Ars Technica. Facts, quotes and full context are in the original article below.