It seems like the news is also coming out that Claude was used in planning the Venezuela and Iran operations[0].
Whatever you thoughts on legality or justification, both have been staggering effective in initial goals.
Interms of intel and complex operation planning.
Despite Trump ragging on Anthropic just hours before it was still considered necessary for the operation.
Quite telling. I don't think Pentagon will be satisfied with just Grok and ChatGPT. Which means OpenAI might really be done for in terms of having no moat beyond least ethics (Grok does that anyways).
[0] https://www.wsj.com/livecoverage/iran-strikes-2026/card/u-s-...