The Pentagon said Friday it signed new agreements with Nvidia, Microsoft, Amazon Web Services, and Reflection to deploy advanced AI capabilities on classified military networks. The move expands a roster that already included OpenAI, Google, SpaceX, and Oracle, giving the department eight named frontier AI vendors tied to its higher-security environments.

What changed

In its release, the department said the companies will help deploy AI capabilities into Impact Level 6 and Impact Level 7 environments, which are used for highly sensitive national security workloads. The agency framed the expansion as part of an anti-lock-in strategy, saying it wants long-term flexibility rather than dependence on a single model provider or cloud stack.

TechCrunch reported the new agreements follow earlier Pentagon arrangements with Google, SpaceX, and OpenAI. The department also said more than 1.3 million personnel have already used GenAI.mil, its secure enterprise AI platform, generating tens of millions of prompts and hundreds of thousands of agents in roughly five months.

Why it matters

The notable shift is that frontier AI vendors are no longer being limited to unclassified productivity pilots. They are being brought deeper into classified infrastructure where procurement, oversight, and safety constraints matter much more.

A multi-vendor approach could reduce switching costs and keep the Pentagon from getting trapped inside one company's stack. But it also shows how quickly classified AI access is becoming a competitive battleground for major model labs, cloud platforms, and AI hardware providers.