Versions :<123456789101112Live>
Snapshot 11:Fri, Feb 27, 2026 2:03:46 PM GMT last edited by Vandita

Anthropic Rejects Pentagon's 'Final Offer'

Anthropic MustRejects ReconsiderPentagon's AI'Final Safeguards by Friday, Demands HegsethOffer'

Anthropic Rejects Pentagon's 'Final Offer'
Above: Pete Hegseth in Louisville, Colorado, on Feb. 23. Image credit: AAron Ontiveroz/The Denver Post/Getty Images

The Spin

Anthropic's ideological resistance to military AI applications threatens national security and warfighter effectiveness. The company's refusal to allow unrestricted lawful use of Claude — the only AI model on classified military systems — forces the Pentagon to consider blacklisting them as a supply chain risk. This safety-obsessed stance, rooted in ties to Democratic donors and "AI doomer" philosophy, prioritizes corporate ideology over defending America.

Establishing guardrails against mass surveillance of Americans and fully autonomous weapons is required for responsible AI governance. Anthropic's insistence on preventing catastrophic misuse of frontier AI technology reflects legitimate concerns about democratic governments turning powerful surveillance tools against their own citizens. The Pentagon's demand for unlimited "lawful use" ignores that existing legislation hasn't caught up to AI's unprecedented capabilities.

Metaculus Prediction

There is a 20% chance that the U.S. government will take control of any U.S. AI company or project before 2029, according to the Metaculus prediction community.


The Controversies



Go Deeper


Articles on this story



© 2026 Improve the News Foundation. All rights reserved.Version 6.18.0

© 2026 Improve the News Foundation.

All rights reserved.

Version 6.18.0