Pentagon–Anthropic Rift Deepens Over Military AI

A sharp confrontation has emerged between the U.S. Department of Defense and Anthropic, one of the world’s leading AI labs. On February 27, 2026, Defense Secretary Pete Hegseth designated Anthropic a “supply chain risk,” escalating tensions after stalled negotiations over military deployment of its Claude models.

At the heart of the dispute are contractual “exceptions” sought by Anthropic. The company refused to permit use of its AI systems for mass domestic surveillance of American citizens or for fully autonomous weapons capable of selecting and engaging targets without human oversight.

Anthropic argues these redlines are grounded in democratic values and technical caution. The firm maintains that current AI systems lack the maturity and reliability to safely support such high-stakes missions without meaningful human control.

The Pentagon, however, contends that military AI models must operate without restrictive usage policies or ideological constraints. Officials argue that operational effectiveness requires systems capable of supporting lawful combat scenarios without pre-embedded limitations.

President Trump further intensified the standoff by directing federal agencies to phase out Anthropic technologies within six months. Combined with the “supply chain risk” designation under 10 USC 3252, the move effectively restricts defense contractors from commercial engagement with the company.

The technology sector is now polarized. OpenAI has reportedly expanded collaboration within classified defense networks, while hundreds of employees across major tech firms have voiced support for Anthropic’s ethical stance through open letters.

This confrontation signals a potential structural decoupling between frontier AI labs and the state. The broader question now facing the industry is profound: will next-generation AI systems be shaped primarily by corporate ethical frameworks—or by the unrestricted strategic imperatives of national defense?