On May 1, 2026, the Pentagon signed deals with AWS, Google, Microsoft, NVIDIA, OpenAI, SpaceX, Reflection, and Oracle to run their AI models on Impact Level 6 and 7 networks — the Secret and Top Secret tiers. Anthropic, which most observers assumed was a lock, didn’t make the list.
What the deals cover
Eight vendors get to deploy their frontier models directly inside the most sensitive U.S. military networks for data fusion, battlefield decision support, and situational awareness. IL6 covers Secret-level workloads, IL7 covers Top Secret. Until now no commercial AI vendor had production clearance for either. This is the moment frontier models cross from civilian infrastructure into the classified stack.
Why Anthropic got cut
Anthropic refused to remove the usage-policy clauses banning domestic mass surveillance and fully autonomous weapons. The Pentagon wanted them out. Anthropic wouldn’t move, and the contract walked. Awkward subplot: the NSA had quietly been running Anthropic’s unreleased Mythos model already. That arrangement now looks fragile.
The split is the real story. The AI defense market just bifurcated into vendors that ship whatever the buyer asks for and one company drawing a line in the sand. Whoever wins this fight long-term ends up defining what an “AI lab” is allowed to mean.
You Might Also Like
- 13b Into Openai yet Microsoft Copilot Cowork Runs on Anthropic Claude
- The Anthropic vs Pentagon Standoff is the ai Story Nobody can Look Away From
- Google is Deploying Gemini ai Agents to 3 Million Pentagon Employees Heres What That Means
- Cursor Composer 2 Takes on Anthropic and Openai With a 0 50 m Token Coding Model and the Benchmarks Back it up
- Openai Burned 15m a day on Sora Google Vids 2 0 is Giving ai Video Away for Free
Top AI Product 
Leave a comment