US District Judge Rita Lin ruled on Thursday that the Trump administration's designation of Anthropic as a supply-chain risk to national security was "unlawful retaliation" violating the First Amendment and was "arbitrary and capricious." The label had been imposed after Anthropic refused to remove safety guardrails that would have permitted Claude to be deployed for fully autonomous weapons systems and mass surveillance of American citizens. The Pentagon had simultaneously sought to designate Anthropic as essential to national security whilst pursuing contracts with the company — a contradiction the judge found legally untenable.

The ruling establishes a precedent that extends beyond Anthropic's specific circumstances. AI safety guardrails — the policies and technical constraints that limit what an AI model can be instructed to do — are now legally established as a form of protected corporate speech rather than removable compliance overhead that government procurement requirements can demand be stripped. For the broader AI industry operating under increasing governmental pressure to expand capability deployments, the ruling provides a defensible legal position: a company can decline to disable its safety policies without accepting that its market position can be lawfully penalised as a result. The durability of that position will depend on whether it survives appeal, and on whether the specific First Amendment framing Judge Lin applied holds in subsequent cases with different factual patterns. For now, it is the most significant legal protection for AI safety constraints that has been established by US courts.