US Military Reportedly Used Anthropic’s Claude in Iran Strikes Despite Trump Ban
Multiple international news organizations have reported that the United States military used the artificial intelligence system Claude, developed by the company Anthropic, during coordinated strikes on Iran — even though President Donald Trump had ordered federal agencies to halt use of the company’s AI tools just hours earlier.
The reports highlight a growing conflict between the U.S. government and AI companies over military applications of artificial intelligence.
Where the Reports Came From
The information was reported by:
- The Guardian
- The Wall Street Journal
- Axios
- The New York Times
- i24 News
According to these outlets, Claude — Anthropic’s AI model — was used during a joint U.S.–Israel operation targeting Iranian leadership in Tehran.
How Claude Was Used in the Operation
Reports state that Claude was used by U.S. military command and defense systems for:
- Intelligence analysis
- Target identification
- Battlefield simulations
- Combat scenario planning
- Intelligence assessments
The Wall Street Journal reported that U.S. military headquarters, including United States Central Command, were using Claude for operational planning purposes.
The New York Times further reported that the Central Intelligence Agency had been tracking Iranian Supreme Leader Ali Khamenei for months. The CIA reportedly gathered high-confidence intelligence about his movements and learned of a rare Saturday morning meeting of senior Iranian officials in central Tehran.
According to the Times, the timing of the strike was adjusted to coincide with this meeting.
Trump’s Order to Sever Ties With Anthropic
Just hours before the reported Iran operation began, President Trump announced that all federal agencies must immediately stop using Anthropic’s AI tools.
On Truth Social, Trump described Anthropic as a “Radical Left AI company run by people who have no idea what the real World is all about.”
The dispute reportedly escalated after Anthropic objected to the military’s use of Claude in a previous January operation involving Venezuelan president Nicolás Maduro. Anthropic cited its terms of service, which prohibit the use of its AI tools for violent purposes, weapons development, or surveillance.
Defense Secretary Pete Hegseth publicly criticized Anthropic, accusing the company of “arrogance and betrayal.” He stated that U.S. warfighters “will never be held hostage by the ideological whims of Big Tech.”
However, despite Trump’s order, the Pentagon reportedly allowed a transition period of up to six months to phase out the technology because Claude was already deeply embedded in defense systems.
CIA Intelligence and the Strike on Tehran
According to reporting by The New York Times:
- The CIA gained increasing confidence about Khamenei’s location patterns.
- Intelligence indicated a rare leadership meeting in central Tehran.
- U.S. intelligence was shared with Israel.
- The timing of the attack was shifted from night to daytime to take advantage of the leadership gathering.
The Times reported that several senior Iranian officials were present at the meeting. The operation was described by an Israeli security official as achieving “tactical surprise.”
OpenAI Steps In
Following the dispute with Anthropic, rival AI company OpenAI reportedly reached an agreement with the Pentagon to provide AI tools for use in classified defense networks.
OpenAI CEO Sam Altman confirmed that an arrangement had been made for Pentagon use of its systems, including ChatGPT.