Anthropic wins first court battle: judge rules Pentagon's supply-chain risk label was unlawful retaliation

The full story goes back to earlier this year. Per TechCrunch, Hegseth and President Trump labeled Anthropic a supply-chain risk and ordered all federal agencies — even those outside the Department of Defense — to stop working with the Claude maker. For an AI company dependent on government contracts and federal market trust, it amounted to a commercial 'ban order.' The trigger was Anthropic's insistence on safety guardrails in defense cooperation talks: it explicitly refused to allow the Pentagon to use its models for fully autonomous weapons and mass surveillance of American citizens. The Pentagon denied it would use the models for anything but lawful purposes. The disagreement escalated from business negotiations into legal warfare.

Judge Rita Lin's ruling was unusually sharp. Per TechCrunch, Lin said the government's 'words and deeds confirm that the challenged actions were based on a desire to make a public example out of Anthropic for its 'arrogance' in criticizing the government.' She highlighted the contradiction between the supply-chain label and other government actions — for instance, Hegseth's proposal to apply the Defense Production Act to Anthropic, which would mean the company was essential to national security rather than a threat to it. She also noted that Anthropic 'undisputedly lacks' any backdoor access to its technology once handed over to the DOD. In other words, calling it a threat while designating it a critical supplier — the government's own stance was incoherent.

The most consequential line in the ruling is Judge Lin's warning about the 'national security' rationale itself: 'Though the Department of War is undisputedly free to select the AI vendor of its choice, the evidence demonstrates that the broad measures imposed on Anthropic were illegal and baseless. The empty invocation of national security is not a blank check to punish and retaliate against government critics.' The significance extends far beyond Anthropic's win or loss — it draws a judicial boundary around whether the government can use national security as cover to punish tech companies that dare to criticize it. As the AI industry grows increasingly dependent on government contracts while colliding with the government over safety and content regulation, this ruling provides an important legal precedent for other companies facing similar pressure.

For the AI industry, the deeper meaning of this case is the increasingly sharp structural tension: frontier AI companies want both government contracts — defense, intelligence and public sector are enormous markets — and the right to hold their safety lines — no fully autonomous weapons, no mass surveillance of citizens. Anthropic's position is 'both': take government money but reserve the right to say no. The government's response was 'either you follow us completely, or you don't do business with us.' This time the court sided with the company, but the game is far from over — the Pentagon will not easily give up access to the most advanced AI capabilities, and Anthropic must keep balancing its safety stance against commercial interests. Expect the debate over 'the boundaries of AI-company government cooperation' to become one of the defining AI policy issues of the coming years.