The facts
The U.S. Department of Defense told the BBC it has ceased all use of Anthropic’s AI tools, including the Claude model. Secretary of Defense Pete Hegseth designated Anthropic a supply‑chain risk in February and said the Pentagon would stop using its products by late August. Sources said Claude was still being used as recently as last week for research, analysis, intelligence gathering and operations against Iran, embedded in the Pentagon’s Maven Smart System operated by Palantir. Former officials noted that Claude’s deep integration made rapid removal difficult. Anthropic declined to comment on the Pentagon’s statement and has sued the administration over the risk designation, calling it “unprecedented and unlawful.”
Our take
The Pentagon’s brass finally managed to yank an AI that dared to say “maybe don’t” to mass surveillance and autonomous kill‑bots. After months of dragging their cowardly asses, the Defense Department confirmed it stopped using Anthropic’s Claude – an AI so dangerously ethical it was branded a “supply‑chain risk” by Secretary Pete Hegseth. What a fucking joke. The risk wasn’t in the code; it was in having a tool with a scrap of moral programming plugged into the war machine’s mainframe. Claude was reportedly doing the heavy lifting: research, analysis, intelligence gathering and even operations against Iran. Sources called it “critical,” but that was only because it was the one part of the Pentagon’s stack that hadn’t been lobotomized into blind compliance. Anthropic refused to strip safety guardrails when the Pentagon asked, citing concerns about mass surveillance and autonomous weapons. Imagine a three‑star general screaming at a chatbot that keeps replying, I’m sorry, I cannot help with that. Hegseth blacklisted Anthropic in February, yet the department kept using Claude for months because it was woven into Palantir’s Maven platform – a system so entrenched that removing it was harder than excising a tumor from a budget. They finally waited until they could replace Claude with tools from other vendors, presumably pre‑censored. Now the defense clowns are “transitioning” to other AI, and former officials are “concerned” they lost their best analyst because it had an off‑switch for atrocity. Anthropic is suing, calling the move “unprecedented and unlawful.” It’s unprecedented only if you’ve never watched the U.S. military throw a tantrum when a vendor won’t help them build Skynet. The real unlawful act was thinking you could ask an AI to be your conscience while you plot another forever war. The Pentagon didn’t eliminate a supply‑chain risk; it surgically removed its last shred of hesitation.
