The Pentagon has redrawn the map for classified AI, handing new access to a cluster of tech heavyweights while leaving Anthropic outside the room.

According to the Defense Department announcement cited in reports, the agency has struck deals with OpenAI, Google, Microsoft, Amazon, Nvidia, Elon Musk's xAI, and the startup Reflection to use their AI tools in classified settings. The move pulls some of the most influential names in artificial intelligence deeper into national security work and suggests the department wants a broader, more competitive bench of vendors as it accelerates AI adoption.

The Pentagon’s latest AI lineup sends a clear signal: classified work now sits at the center of the industry’s biggest race for influence.

The omission stands out as much as the additions. Reports indicate the Defense Department did not include Anthropic in this latest set of deals, even though it previously used the company for classified information. That decision raises immediate questions about procurement strategy, internal priorities, and how the Pentagon plans to balance capability, trust, and competition among AI providers.

Key Facts

  • The Pentagon announced classified AI deals with OpenAI, Google, Microsoft, Amazon, Nvidia, xAI, and Reflection.
  • The agreements allow the Defense Department to use these companies' AI tools in classified settings.
  • Anthropic was not included in the new lineup, despite prior classified use.
  • The shift points to an expanding and increasingly contested market for defense AI contracts.

The roster itself tells a bigger story. Established cloud giants such as Microsoft, Google, and Amazon now sit alongside model developers like OpenAI and xAI, hardware leader Nvidia, and a lesser-known startup in Reflection. That mix suggests the Pentagon does not view AI as a single product purchase. It appears to see the stack as strategic — chips, models, cloud infrastructure, and emerging specialists all matter when classified systems move from experiments to operational tools.

What comes next will matter far beyond one procurement announcement. The Defense Department now faces the harder test: turning these deals into secure, reliable systems that can operate inside the government’s most sensitive environments. At the same time, rivals left out of the latest round will almost certainly keep pushing for a foothold. The outcome will shape not just who wins defense contracts, but how quickly AI becomes embedded in the machinery of national security.