Anthropic gets its first court win over the Pentagon’s supply chain risk label | TechCrunch

A federal district judge in California has issued a ruling declaring that the United States government's designation of artificial intelligence developer Anthropic as a national security supply chain threat was unlawful.

District Judge Rita Lin determined that the label, instituted by the Trump administration, represented an act of illegal retaliation against the startup, violating both its First Amendment rights and its Fifth Amendment right to due process. The high-profile conflict between the technology firm and federal officials originally stemmed from Anthropic’s decision to implement strict safety guardrails on its Claude language models, which prevented defense agencies from utilizing the software for fully autonomous warfare and domestic mass surveillance programs.

In her written opinion, Judge Lin characterized the administration's blacklisting of the artificial intelligence startup as an arbitrary effort to penalize the company for speaking out against government demands. The court pointed out significant contradictions in the government's actions, observing that defense officials simultaneously attempted to acquire Anthropic's services under the Defense Production Act and collaborated with the company on cybersecurity projects while publicly labeling them a threat. While the judge acknowledged that defense agencies are free to select their preferred software vendors, she emphasized that national security concerns cannot be used as a pretext to retaliate against critics. Anthropic welcomed the decision, though a separate legal challenge filed by the company in Washington, D.C., is still pending resolution.