A federal judge has put a stop to the Pentagon's attempt to label AI company Anthropic a national security risk, overturning a decision that would have barred it from federal contracts. The judge ruled the move was unconstitutional retaliation.
US District Judge Rita Lin in California struck down Defense Secretary Pete Hegseth's February decision that classified Anthropic as a "supply-chain risk." This designation not only made the company ineligible for federal contracts but also penalized contractors and suppliers from doing business with Anthropic. Judge Lin found Hegseth's actions to be "arbitrary, capricious, an abuse of discretion, and otherwise not in accordance with law."
The judge stated that while the Department of War is free to choose its AI partners, the measures taken against Anthropic were illegal and unfounded. Lin also determined that nine government agencies, including the Pentagon, Treasury, State Department, and Homeland Security, had improperly sanctioned Anthropic, and those penalties are now lifted.
However, the ruling doesn't force the Pentagon to use Anthropic's AI models, nor does it prevent the department from choosing other options. While the Pentagon hasn't commented, an appeal is expected. Anthropic expressed its pleasure with the court's decision, emphasizing its commitment to working with the government on AI for national security.
This legal battle stems from a disagreement earlier this year over a $200 million deal. Anthropic's AI models, like Claude, were set to be used for military applications, but tensions flared after reports suggested they were used in an operation to capture Venezuelan president Nicolás Maduro. Concerns were raised by an Anthropic staffer via a Palantir employee regarding the model's deployment.
Anthropic pushed for limitations on how its technology could be used, specifically prohibiting support for lethal autonomous weapons and mass surveillance. Hegseth rejected these curbs, arguing that contractors shouldn't dictate post-deployment usage and insisting on "all lawful use." When negotiations failed, Hegseth's designation of Anthropic as a "supply-chain risk" was widely viewed as punitive, effectively blacklisting the company from federal business.
The Pentagon had defended the designation, citing unacceptable risks if Anthropic could potentially disable or alter its technology, especially during wartime. Anthropic responded by filing lawsuits, alleging violations of its First and Fifth Amendment rights. The judge noted that the government is currently in discussions with Anthropic about collaborating on its new "Mythos" model, which contradicts the idea that Anthropic poses a genuine national security threat.
