A United States federal judge has blocked the Department of Defense from designating AI developer Anthropic as a national security supply-chain risk, ruling that the Pentagon's blacklisting action was unlawful and unsupported by evidence.
In a 59-page decision, U.S. District Judge Rita Lin of the Northern District of California found that the defense agency overstepped its statutory authority when Defense Secretary Pete Hegseth designated Anthropic under a procurement statute originally designed to protect military infrastructure from foreign sabotage. Judge Lin concluded that the government's designation lacked factual basis and penalized the company for raising public debate regarding AI safety standards and defense contracting terms.
Legal Context and Contracting Dispute
The dispute originated earlier in 2026 following negotiations over the operational deployment of Claude models across defense networks. Anthropic maintained contractual restrictions barring the use of its foundation models in autonomous kinetic weapons systems and domestic mass surveillance operations. In response, the Department of Defense initiated supply-chain risk designations intended to restrict defense contractors from utilizing Anthropic software.
Anthropic filed suit in federal court in March 2026, challenging the administrative designation as arbitrary and retaliatory. In her ruling, Judge Lin noted that penalizing an AI vendor for seeking contractual limits on military applications violated procedural standards under federal administrative law.
Anthropic issued a statement welcoming the court's decision, emphasizing that the legal challenge was necessary to ensure predictability for commercial and public sector partners while maintaining commitment to safe frontier AI deployment.
