7 of the biggest names in AI, including OpenAI, Google, Microsoft, and Amazon, are now moving into military systems — but one major player is missing from the list, and that raises uncomfortable questions about where this technology is heading

The U.S. Department of Defense has signed deals with OpenAI, Google, Microsoft, Amazon Web Services, Nvidia, SpaceX, and Reflection AI to integrate their AI systems into military operations. The partnerships fall under the Pentagon’s “AI-first fighting force” initiative, with tools approved for any “lawful use”—a broad standard that covers data processing, decision support, and potentially autonomous systems. The companies bring distinct strengths: chips from Nvidia, cloud from AWS, and frontier models from OpenAI and Google. Tens of billions of dollars are earmarked for these programs, reflecting a strategic push to embed AI into the infrastructure of modern warfare.

Conspicuously absent is Anthropic, developer of the Claude model. The company has been negotiating with the DoD for months but refuses to allow its AI for domestic surveillance or fully autonomous lethal weapons, pointing to risks like models tolerating nuclear threats in simulations. In response, the Pentagon labeled Anthropic a supply-chain risk to block it from contracts, a move the company is challenging. While other firms accept broad terms for access and influence, Anthropic’s holdout exposes unresolved tensions over oversight, accountability, and unintended consequences as military AI deployment accelerates.