A federal judge on July 30 said the Defense Department (DOD) has not produced sufficient evidence to justify its designation of Anthropic as a supply-chain risk, as both parties seek summary judgment in the AI company’s legal challenge. At a hearing in San Francisco, Judge Lin questioned the government’s assertion that Anthropic posed a risk because it could modify its AI models after deployment, stating she saw no evidence that Anthropic could alter the model after delivery or flip a kill switch. The case tests the government’s authority to restrict AI providers from federal networks on national security grounds. During a March hearing, Lin called the government’s treatment of Anthropic “troubling” and later issued a preliminary injunction blocking enforcement of the designation, finding Anthropic demonstrated a likelihood of success on the merits. The ruling did not require the DOD to retain Anthropic technology, and the department has continued removing its software. The dispute stems from disagreements over DOD’s intended use of Anthropic’s Claude models for “all lawful purposes,” including sensitive military applications, while Anthropic objected to mass surveillance and fully autonomous weapons. After President Donald Trump directed agencies to sever ties, Defense Secretary Pete Hegseth designated the company a supply-chain risk, and Anthropic filed two lawsuits alleging unlawful retaliation. Parallel proceedings in the D.C. Circuit could send the dispute to the U.S. Supreme Court. Under the president’s directive, defense components must remove all Anthropic software by Sept.