Judge Rita Lin of the Northern District of California issued a preliminary injunction last Wednesday blocking the Pentagon from designating Anthropic as a supply chain risk, a ruling that legal experts called devastating for the government's case but that has not, in practice, restored the AI company's access to federal contracts [1]. The government has seven days to appeal, and the Trump administration has given every indication it intends to do so.

As we reported in our previous coverage, Anthropic filed two federal lawsuits on March 9 after the Pentagon branded it a security threat and President Trump signed a directive banning federal agencies from purchasing Claude, the company's AI model [2]. The designation came after Anthropic publicly criticized the Pentagon's approach to AI procurement safety standards, which the company described as inadequate.

Judge Lin's ruling was unambiguous. The court found that the Pentagon's own records showed the supply chain risk designation was imposed in retaliation for Anthropic's public criticism, not because of any genuine security concern [3]. The judge wrote that the government failed to identify a single instance of Claude being used to compromise federal systems or data. Fortune described the case as "attempted corporate murder" [4].

The tech industry responded with cautious optimism. The Computer and Communications Industry Association welcomed the ruling, and Microsoft -- which had quietly signaled its concern about the precedent -- indicated it would file an amicus brief if the case reaches the appellate level [5]. The Atlantic Council published an analysis warning that the broader crisis of trust between the tech sector and the defense establishment could hamper AI development for national security purposes regardless of the legal outcome [6].

But Anthropic's real problem is practical, not legal. Even with the injunction in place, no federal agency has resumed procurement discussions with the company. The label, while technically blocked, has created a chilling effect that extends beyond government contracts to enterprise customers who fear association with a company the Pentagon targeted.

CNBC reported that Judge Lin pressed DOD lawyers during oral arguments on March 24 to explain why Anthropic was singled out, receiving what she characterized as "circular" answers. The government's theory -- that Anthropic's public advocacy on AI safety somehow constituted a supply chain risk -- was described by Lawfare as "bypassing the ordinary procurement process".

The seven-day stay expires this week. If the administration appeals, the case moves to the Ninth Circuit, where it could take months to resolve. In the meantime, Anthropic remains in limbo: legally vindicated, commercially damaged, and waiting.

-- DAVID CHEN, San Francisco