TL;DR
The US military reportedly used Anthropic’s Claude AI for intelligence and target selection during the joint US-Israel bombardment of Iran — just hours after President Trump ordered all federal agencies to sever ties with the company. The situation highlights how deeply embedded AI tools have become in military operations.
AI in the firing line
According to the Wall Street Journal and Axios, US military command used Claude for intelligence purposes, target selection, and battlefield simulations during the massive bombardment of Iran that began on Saturday.
The timing is striking. On Friday, Trump had ordered all federal agencies to stop using Claude immediately, denouncing Anthropic on Truth Social as a “Radical Left AI company run by people who have no idea what the real World is all about.”
How the rift developed
The dispute traces back to January, when the US military used Claude during a raid to capture Venezuelan president Nicolás Maduro. Anthropic objected, citing its terms of use which prohibit Claude from being applied for violent ends, weapons development, or surveillance.
Relations deteriorated from there. Defense Secretary Pete Hegseth accused Anthropic of “arrogance and betrayal” in a post on X, adding that “America’s warfighters will never be held hostage by the ideological whims of Big Tech.” He demanded full, unrestricted access to all Anthropic models for every lawful purpose.
Yet Hegseth also acknowledged the difficulty of rapidly decoupling military systems from Claude, given how widely it is now used. He said Anthropic would continue providing services “for a period of no more than six months to allow for a seamless transition.”
OpenAI steps in
Since the break with Anthropic, rival company OpenAI has moved quickly to fill the gap. CEO Sam Altman confirmed he had reached agreement with the Pentagon for use of OpenAI’s tools, including ChatGPT, within its classified network.
Looking forward
The episode raises fundamental questions about what happens when AI companies’ ethical commitments conflict with government demands — and whether critical military infrastructure can realistically be decoupled from AI systems already woven into operations. The six-month transition period suggests the Pentagon itself recognises this is not a simple switch to flip.