Judge Finds US Defense Department Unlawful in Designation of AI Startup Anthropic

The federal courts have delivered a decision that upends a high‑profile move by the Pentagon. Judge Rita Lin of the US District Court declared that the Department of Defense unlawfully singled out artificial‑intelligence startup Anthropic as a supply‑chain risk—an action the defence ministry had said was to protect national security.

Anthropic, which developed the Claude model, had publicly stated that its AI cannot safely be used in weapons systems or for domestic surveillance because of risks to privacy and safety. The company’s refusal to accept Pentagon‑issued contract terms triggered an administrative inquiry that culminated in a public designation that effectively barred the company from doing business with military contractors.

Lin warned that "citing national security is not a blank check to punish and retaliate against government critics." The decision suggests that the Pentagon cannot arbitrarily use the supply‑chain‑risk label to silence dissent. The judge highlighted that the government’s actions appear to be "classic First Amendment retaliation."

The ruling was welcomed by Anthropic, whose spokesperson said the company aims "to work productively with the government to harness AI for national security so all Americans benefit from this technology."

The case illustrates a mounting tension between US civil liberties—particularly the freedom of speech—and the government’s desire to control military use of emerging technology. It also signals a potential shift in how federal agencies can designate private firms as threats, a practice that has so far been reserved for entities in adversarial nations.

The Pentagon’s designation of Anthropic as a supply‑chain risk marked the first time an American company received this public label. Future legal challenges may rely on this precedent to argue that national‑security justifications cannot override basic constitutional protections.