What happened
On August 27, 2026, U.S. District Judge Rita Lin (N.D. Cal.) issued a 59-page ruling in Anthropic PBC v. U.S. Department of War, finding that the Department of Defense's February 2026 designation of Anthropic as a 'supply chain risk' — which barred all federal agencies and DoD contractors from doing business with the company — was unlawful retaliation in violation of the First Amendment, violated the Fifth Amendment Due Process Clause, and was 'arbitrary, capricious, and in excess of statutory jurisdiction' under the APA. The court permanently enjoined DoD from enforcing the designation and ordered rescission of related guidance. The dispute arose after Anthropic refused Pentagon demands to remove Claude usage-policy safeguards against fully autonomous weapons and mass domestic surveillance.
Why it matters
This is the first-ever legal test of the government's 'supply chain risk' national-security procurement tool being used against a major US AI lab over its safety policies, and the court's ruling constrains the government's ability to retaliate against AI vendors for refusing to remove safety guardrails or for publicly criticizing government AI policy. It sets an important precedent limiting how national-security tools can be wielded against frontier AI developers, directly shaping the relationship between AI labs and government procurement/national-security policy. A second related Anthropic suit is still pending in D.C. federal court.
Action needed
AI labs and defense contractors should monitor whether DoD appeals or reissues a narrower supply-chain-risk designation; contractors previously barred from working with Anthropic should reassess compliance posture given the injunction.