Pentagon Secures AI Access on Classified Networks from Seven Major Vendors

TL;DR: The Pentagon has signed contracts with seven leading AI firms to run their models on classified networks, expanding the GenAI.mil platform used by 1.3 million personnel.

Context The Department of Defense announced Friday that it will integrate AI capabilities from Amazon Web Services, Google, OpenAI, Microsoft, Nvidia, Reflection and SpaceX into its secure, classified infrastructure. The move follows a White House directive that halted the use of Anthropic’s products after the company refused to allow unrestricted military use.

Key Facts - Seven AI providers will now supply models that can operate on the Pentagon’s classified networks, a step the department describes as essential for “lawful operational use.” - The contracts stress avoiding vendor lock‑in, a phrase that means the military wants the ability to switch providers without costly re‑engineering. - GenAI.mil, the DoD’s internal AI platform, has logged 1.3 million users generating tens of millions of prompts and deploying hundreds of thousands of AI agents in just five months. - The platform already hosts Google Cloud’s Gemini for Government AI, supporting unclassified processes for 3 million civilian and military staff. - Officials say the new tools will speed data synthesis, improve situational awareness and augment decision‑making for warfighting, intelligence and enterprise missions.

What It Means By spreading AI workloads across a “resilient American technology stack,” the Pentagon aims to give warfighters rapid, reliable access to cutting‑edge models while retaining the freedom to replace or add vendors as needs evolve. The emphasis on a diverse suite of capabilities reflects concerns that reliance on a single supplier could create strategic vulnerabilities. The expanded use of GenAI.mil suggests the department is moving from experimental pilots to routine operational integration, cutting task timelines from months to days.

The agreements also signal a broader government push to keep advanced AI development within U.S. borders, reinforcing a domestic ecosystem that can meet national‑security demands. As the platform scales, the next focus will be on monitoring performance, security compliance and the ethical boundaries of AI‑enabled warfare.

What to watch next: Implementation details, including which specific models will be deployed and how the Pentagon will measure operational impact, are expected in the coming weeks.