Pentagon Signs Classified AI Deals with 8 Tech Firms, Excludes Anthropic

The United States Department of Defense isn't just experimenting with artificial intelligence anymore; they are embedding it deeply into their most secure operational networks. In a sweeping move, the Pentagon recently finalised agreements with eight major technology providers: Google, OpenAI, Nvidia, SpaceX, Microsoft, AWS, Reflection AI, and Oracle. The goal is to deploy frontier AI models on classified Impact Level 6 (IL6) and Impact Level 7 (IL7) networks.

These are the highly restricted environments where sensitive intelligence, operational planning, and national security data reside. The initiative is designed to drastically accelerate data synthesis and improve situational awareness for warfighters across all domains.

This isn't a small pilot programme. The scale of AI adoption within the military is already staggering. Currently, over 1.3 million personnel are using GenAI.mil, the Pentagon's secure enterprise platform. While that platform primarily handles unclassified tasks like document drafting and basic research, moving these capabilities to IL6 and IL7 networks changes the game entirely. It means these models will now be processing classified intelligence intercepts, satellite imagery analysis, and strategic logistics data.

The Pentagon has explicitly stated that these agreements are meant to accelerate the transformation toward establishing the United States military as an "AI-first fighting force."