In a landmark legal ruling that sets a new precedent for tensions between private artificial intelligence developers and the U.S. military, a federal district judge has found that the U.S. Department of Defense acted unlawfully when it labeled leading AI startup Anthropic a national supply chain security risk, deeming the agency’s action “illegal and baseless”.
The clash between the two parties stems from Anthropic, the developer of the widely used large language model Claude, declining the Pentagon’s request to grant the U.S. military access to its proprietary AI models for use cases including mass surveillance and the development of autonomous weapons systems. Refusing to comply with the military’s demand, the AI developer argued that its technology was not sufficiently tested or reliable to be deployed safely in weapons systems. It also reiterated its longstanding opposition to using AI for domestic surveillance, framing such use as a violation of fundamental civil rights.
Following the company’s refusal, the Pentagon publicly designated Anthropic as a supply chain risk under a statute originally crafted to shield U.S. military infrastructure from foreign sabotage. This marked the first time in U.S. history that a domestic American tech company received such a public, official designation. The Biden administration doubled down on the Pentagon’s position, with White House officials publicly attacking Anthropic as “a radical left, woke company” that was overstepping by attempting to restrict U.S. military operations. The Pentagon argued at the time that private sector firms have no right to place constraints on U.S. military action, asserting that the armed forces answer only to the U.S. Constitution, not a private tech company’s terms of service.
Anthropic swiftly filed a lawsuit challenging the designation, calling the government’s actions both “unprecedented and unlawful”, and argued that Defense Secretary Pete Hegeth had abused his authority by imposing the label in retaliation for the company’s policy choices. In her ruling, U.S. District Judge Rita Lin rejected the government’s national security justification, writing that “citing national security is not a blank check to punish and retaliate against government critics”.
The designation carried severe potential consequences for Anthropic: company executives have projected that the label could cost the startup billions of dollars in lost commercial partnerships and irreversible reputational damage. As of Wednesday, the BBC has confirmed it has reached out to the Pentagon, the White House and Anthropic to request official comment on the ruling, with no formal responses released publicly at the time of reporting.