Anthropic fights DOD’s ‘unacceptable risk’ designation in court

The U.S. Department of Defense (DOD) has formally labeled Anthropic an “unacceptable risk to national security” in a 40-page court filing submitted on Tuesday evening. The filing, lodged in a California federal court, represents the agency’s first rebuttal to the AI company’s lawsuit. The legal dispute challenges a decision made last month by Defense Secretary Pete Hegseth to label Anthropic a “supply-chain risk.” As part of its legal action, Anthropic has requested that the court temporarily block the DOD from enforcing this designation.

The conflict stems from a $200 million contract signed last summer between Anthropic and the Pentagon. During subsequent negotiations over the contract’s terms, Anthropic established corporate “red lines”—defined as corporate policy constraints or non-negotiable terms. Specifically, the company stated it does not want its AI systems used for mass surveillance or lethal weapons targeting. In response, the DOD argued in its filing that Anthropic might attempt to disable or alter its AI models during warfighting operations if these corporate red lines are crossed. Anthropic CEO Dario Amodei sought to clarify the company’s stance in late February, stating that the company understands the Department of War, not private companies, makes military decisions, and that Anthropic has never raised objections to specific military operations or attempted to limit the use of its technology in an ad hoc manner.

Legal experts have questioned the government’s justification for the designation. Chris Mattei, a lawyer specializing in First Amendment issues and a former Justice Department attorney, stated that there has been no investigation to support the DOD’s concerns regarding Anthropic potentially disabling its models. Mattei argued that the department’s argument fails to explain how Anthropic’s negotiating position rendered it an “adversary.” According to Mattei, “The government is relying completely on conjectural, speculative imaginings to justify a very, very serious legal step they’ve taken against Anthropic.” He further characterized the administration’s conduct as a “retaliatory punishment” for Anthropic’s refusal to agree to the government’s terms.

The legal battle has drawn significant support from the broader technology sector. Major tech firms, including OpenAI, Google, and Microsoft, have filed amicus briefs—legal documents filed by non-litigants to provide information or insight—in support of Anthropic. A hearing on Anthropic’s request for a preliminary injunction—a court order to stop an action before a final decision—is scheduled for next Tuesday.

Why it matters

This case sets a critical precedent for how private AI labs negotiate with government agencies, specifically regarding the limits of corporate “red lines” in military contracts and the definition of supply-chain risks.