Anthropic Stands Firm: Rejects Pentagon's AI Demands for Lethal AI Weapons, Surveillance

Anthropic, led by CEO Dario Amodei, has formally rejected the Department of Defense's demands for unrestricted access to its artificial intelligence systems. This pivotal decision was announced less than 24 hours before the deadline of an ultimatum issued by the Pentagon, marking the climax of intense public discourse, social media engagement, and private negotiations. The core of the dispute revolves around Defense Secretary Pete Hegseth’s push to renegotiate existing military contracts with all AI laboratories. However, Anthropic has steadfastly maintained its two core principles: a firm stance against mass surveillance of American citizens and the development of lethal autonomous weapons, defined as those capable of engaging targets without any human oversight. While OpenAI and xAI reportedly acceded to the new terms, Anthropic's defiance led to CEO Amodei's summons to the White House for a direct meeting with Secretary Hegseth, where the Secretary reportedly issued a stark ultimatum: comply by the close of business on Friday, or face unspecified consequences.

In a public statement released late Thursday, Amodei articulated his company's foundational commitment: “I believe deeply in the existential importance of using AI to defend the United States and other democracies, and to defeat our autocratic adversaries. Anthropic has therefore worked proactively to deploy our models to the Department of War and the intelligence community.”

Amodei further clarified that Anthropic has “never raised objections to particular military operations nor attempted to limit use of our technology in an ad hoc manner.” However, he stressed that in a “narrow set of cases, we believe AI can undermine, rather than defend, democratic values,” specifically citing concerns over mass domestic surveillance and fully autonomous weapons. He distinguished this by noting that “partial autonomous weapons … are vital to the defense of democracy” and acknowledged that fully autonomous weapons might eventually “prove critical for our national defense.” Yet, he firmly stated that “today, frontier AI systems are simply not reliable enough to power fully autonomous weapons,” leaving open the possibility of future cooperation on such systems once their reliability improves.

In response to Anthropic's stance, the Pentagon had reportedly initiated measures to escalate pressure. These included instructing major defense contractors to evaluate their reliance on Anthropic’s Claude AI, a move interpreted as a precursor to potentially labeling the company a “supply chain risk.” This classification, typically reserved for entities posing direct threats to national security, underscores the severity of the Pentagon’s position. Furthermore, reports indicated that the Department of Defense was considering invoking the Defense Production Act, a powerful executive authority, to compel Anthropic's compliance.

Despite these escalating pressures, Amodei reiterated Anthropic’s unwavering resolve in his statement, asserting that the Pentagon’s “threats do not change our position: we cannot in good conscience accede to their request.” He also outlined a contingency plan, stating, “should the Department choose to offboard Anthropic, we will work to enable a smooth transition to another provider, avoiding any disruption to ongoing military planning, operations, or other critical missions. Our models will be available on the expansive terms we have proposed for as long as required.”