You can find a source for it being used here [1] though parent is misrepresenting it either on purpose or by lacking understanding, I stand to my point above, the timeline is

1. the US used it, in venezuela then to prepare Iran

2. it became public through journals reporting it, but they didn't really hide from it afterwards and told explicitely what and who they used, naming anthropic

3. Diego (the CEO) answered by saying this violate terms and the US is not allowed and should stop that immediately, publicly

4. They have a talk, and while concessions are made it blocks not about war with ennemies but about surveillance at home, which Diego refuses to let go

5. Trump reacted by calling Anthropic a "Radical Left AI company run by people who have no idea what the real World is all about"

6. Peter Hegseth annonce they will stop using it but asked for a stop gap deal "Anthropic will continue to provide the Department of War its services for a period of no more than six months to allow for a seamless transition."

7. Open AI annonce a deal with the governement four and a half HOURS later, with totally cool safeguards "We think our agreement has more guardrails than any previous agreement for classified AI deployments, including Anthropic's", journals note that "OpenAI's new Pentagon deal doesn't explicitly prohibit collecting Americans' publicly available information - a sticking point that rival Anthropic says is crucial for ensuring domestic mass surveillance doesn't take place"

8. In totally unrelated news, Anthropic new models get banned for exports in a move the US has never done before, no other company has their models banned

And now

9. Anthropic change the person talking with the gov to Brown

10. After a bit anthropic models start being unbanned

[1] https://www.theguardian.com/technology/2026/mar/01/claude-an...