Pentagon Signs AI Deals with Seven Companies, Formally Excludes Anthropic
View original source →The Department of Defense formalized the exclusion of Anthropic from classified AI contracts on May 1, signing deals with seven competing companies. The exclusion stems from Anthropic's refusal to allow unrestricted Claude deployment for fully autonomous weapons and mass domestic surveillance.
Key Points:
• The Pentagon designated Anthropic a supply-chain risk to U.S. national security in March 2026 after Anthropic refused to remove safety guardrails required for autonomous weapons systems.
• Anthropic filed two federal lawsuits against the Trump administration challenging the designation as unconstitutional and retaliatory.
• A California federal judge blocked the government's initial effort; despite this, the Pentagon moved forward with the seven-company contract roster.
• Until recently, Claude was the only AI model available in the Pentagon's classified network — this exclusion marks a dramatic reversal of that position.
• The seven companies signed include OpenAI, Google, Microsoft, Amazon, NVIDIA, SpaceX, and Reflection — formalizing AI deployment in military classified networks for 'lawful operational use.'
This is the most consequential AI governance story of 2026. Anthropic is paying a steep commercial price for refusing to remove safety guardrails — but the company is treating this as a principled line it will not cross. The legal outcomes will define what governments can and cannot require of AI providers regarding safety commitments.
Why It Matters: This decision transforms AI governance from abstract policy discussion into commercial reality with billion-dollar stakes. Every major AI provider now faces a documented choice: accept government contracts requiring safety guardrail removal, or face exclusion from significant procurement markets.