In brief
A federal judge has declared that the Pentagon’s designation of artificial‑intelligence startup Anthropic as a supply‑chain risk violated the Constitution. The move was aimed at curbing the company’s refusal to allow its models to be used for military surveillance or autonomous weapons. Judge Rita Lin said the government’s action amounted to classic First Amendment retaliation and was illegal. Anthropic welcomed the decision and reiterated its commitment to cooperate with the US for national security while defending its rights.
A federal judge has ruled that the Pentagon’s move to label AI newcomer Anthropic as a supply‑chain risk was unlawful, underscoring the limits of executive power over corporate speech. The decision was prompted by the company’s refusal to allow its AI models to be used for military surveillance or autonomous weapons, a stance the U.S. defense department reportedly tried to counter.
Judge Rita Lin, writing in the 10‑page ruling, stressed that citing national‑security concerns does not grant a blanket license to punishing critics. She described the Pentagon’s order as “illegal and baseless” and warned that the move could chill public debate on the military use of AI.
Anthropic’s spokesperson said the company’s models were “not reliable enough to be safely used in weapons systems” and that the company opposed domestic surveillance use as a violation of rights. The firm has since welcomed the ruling and re‑affirmed its commitment to work with the government on responsible AI for national security.
The court also noted that the Pentagon’s policy, typically reserved for companies in hostile countries, had never before been applied to an American business. The judge’s remarks about “crippling” the company and chilling debate highlight the first instance of a First Amendment case involving AI in a military context.
Reuters – 28 August 2026, 03:57 BST (updated: 1 hour ago)
