US military reportedly used Claude in Iran strikes despite Trump’s ban

Key Conflict: Defiance of a Federal Ban

Despite an eleventh-hour order from Donald Trump to sever all ties with Anthropic, the U.S. military reportedly used Claude to inform a massive joint US-Israel bombardment of Iran. Trump had ordered federal agencies to stop using the tool immediately, labeling Anthropic a “Radical Left AI company.”

How the Military Used the AI

According to reports from the Wall Street Journal and Axios, the Pentagon utilized Claude for:

• Intelligence gathering: Processing data to inform strategic decisions.
• Target selection: Identifying specific locations and assets for the strike.
• Battlefield simulations: Running scenarios to predict engagement outcomes.

The Source of the Friction

The relationship soured following a January mission to capture Venezuelan President Nicolás Maduro. Anthropic objected to its technology being used in that raid, citing its safety terms of service which prohibit the use of Claude for violence, weapon development, or surveillance.

In response, Defense Secretary Pete Hegseth accused the company of “betrayal,” asserting that the American military should not be restricted by the “ideological whims” of tech companies. Hegseth has demanded unrestricted access to Anthropic’s models for all lawful military purposes.

The Transition and OpenAI’s Role

Because Claude is so deeply integrated into current military systems, the Pentagon cannot detach from it instantly. A six-month transition period has been established to move toward what Hegseth described as “more patriotic” services.

During this fallout, OpenAI has moved to fill the gap. CEO Sam Altman reportedly reached an agreement with the Pentagon to integrate OpenAI’s tools, including ChatGPT, into the military’s classified networks.