It looks like Anthropic isn't as radioactive to other defense contractors - for now, and on paper.
In a sharply worded 43-page order issued Thursday, U.S. District Judge Rita F. Lin (Biden) of the Northern District of California granted Anthropic PBC's motion for a preliminary injunction, blocking key punitive measures imposed by the Trump administration after the AI company publicly refused to lift safety restrictions on its Claude model.
Lin minced no words on the supply-chain label - the core of the dispute:
“Nothing in the governing statute supports the Orwellian notion that an American company may be branded a potential adversary and saboteur of the U.S. for expressing disagreement with the government.”
Recall that Anthropic refused to change the user policy for its AI tool Claude to allow the government to use it for what Anthropic described as “mass surveillance” and “fully autonomous weapons.” After they were branded a Supply Chain Risk, they sued on March. 9, calling the government's actions "unprecedented and unlawful."
Lin ruled that the broad measures “do not appear to be directed at the government’s stated national security interests” and instead “appear designed to punish Anthropic.” One amicus brief called the actions “attempted corporate murder”; the judge noted they “might not be murder, but the evidence shows that they would cripple Anthropic.”
The Injunction
Lin's preliminary injunction bars enforcement of three Challenged Actions; Trump ordering the government to immediately stop using Anthropic tech, Hegseth prohibiting government contractors from 'commercial activity' with Anthropic, and the DoW's formal designation of Anthropic as a “supply chain risk” to national security.
The order does not force the Pentagon to start using Anthropic again, nor does it interfere with a planned six-month phase-out for existing systems if done without the broader bans. A separate parallel challenge to one DoW letter (under 41 U.S.C. § 4713) is pending in the D.C. Circuit; that case remains unaffected.
At the March 24 hearing, DoW counsel conceded that portions of the Hegseth Directive had "absolutely no legal effect" on their own and that DoW did not intend to terminate unrelated commercial relationships - yet declined to stipulate to an injunction, citing ongoing "assessment."
“While this case was necessary to protect Anthropic, our customers, and our partners, our focus remains on working productively with the government to ensure all Americans benefit from safe, reliable AI," Anthropic said in a statement.
The order is a preliminary injunction only; the case will proceed to full merits. But Judge Lin’s thorough factual record and legal analysis make clear that branding a domestic AI firm a national-security threat for publicly advocating safety guardrails crosses a bright constitutional line. The government retains full authority to choose its tools—just not to punish a company for speaking out about their limitations.
