Anthropic wins court ruling that Pentagon blacklist was unlawful

Judge Rita F. Lin of the U.S. District Court for the Northern District of California ruled Thursday that the Department of Defense’s decision to place Anthropic on a supply chain risk list was unconstitutional, describing it as prohibited retaliation that violates the First Amendment. The decision ends a months-long dispute that began when the lab refused to sign a new contract allowing the military unrestricted use of its technology.

The conflict started last winter when Defense Secretary Pete Hegseth demanded that all artificial-intelligence labs reopen existing contracts and agree to a clause permitting AI use for “any lawful purpose,” a major expansion of Pentagon authority. Most companies, including Google, Microsoft, OpenAI and SpaceX, eventually signed. Anthropic insisted on two safeguards: a ban on mass surveillance of U.S. citizens and a ban on lethal autonomous weapons—AI systems that make kill decisions without human oversight.

Anthropic’s refusal set off a chain of reciprocal pressure that culminated in a government ultimatum. Less than 24 hours before the deadline, CEO Dario Amodei issued a statement that the company “has never opposed specific military operations” but “in a small handful of cases we believe AI could undermine, rather than protect, democratic values.” In response, the Pentagon placed the company on the supply chain risk list, a classification normally reserved for national-security threats, and began pursuing contracts with seven alternative labs.

In March, Judge Lin issued a temporary injunction that blocked the blacklist, noting that the Defense Department’s filings showed the classification was based on “adverse conduct through the media.” In the final judgment, Lin wrote that “empty national-security enforcement is not a blank check to punish and retaliate against government critics,” adding that Hegseth’s decision was “arbitrary and vindictive.” She clarified that the military may choose its AI supplier, but the sweeping measures taken against Anthropic were unlawful and unfounded.

In an official release, spokesperson Danielle Ghiglieri said Anthropic welcomes the ruling and will continue to focus on productive work with the government “to harness AI for national security so that all Americans benefit from the technology.” The open question is whether the Pentagon will return to the table with the original terms Anthropic proposed, or will continue with suppliers that have already accepted the expanded conditions.