United States Uses AI in Attack on Iran, Also Employed in Maduro Capture
The United States has involved artificial intelligence technology from Anthropic in recent attacks against Iran. The same AI tools were deployed during US military operations that removed Venezuelan leader Nicolás Maduro in January.
Coordinated waves of US attacks, launched jointly with Israel since Saturday (28 February), naturally involved conventional weapons including Tomahawk cruise missiles, stealth fighter jets, and for the first time, one-way combat drones or kamikaze drones.
However, the use of AI devices from Anthropic attracted particular attention, given that the equipment was prohibited by President Donald Trump mere hours before Washington bombed Tehran and other regions on Saturday (28 February) local time.
Trump described Anthropic as a “radical left-wing AI company run by people who don’t know what the real world is”. The US even designated Anthropic as a supply chain risk, implying the company posed a threat to national security.
According to a source familiar with the situation, as reported by Reuters and The Guardian on Monday (2 March 2026), the Pentagon used AI services from Anthropic, including the Claude device—a type of AI assistant—during its attack on Iran over the weekend.
It remains unclear exactly how the AI equipment was deployed in military operations. Neither the Pentagon nor Anthropic has commented.
Anthropic’s artificial intelligence, from the San Francisco-based AI company, has been used across the intelligence community and armed forces, and was the first among similar AI companies to work with classified information through a supply agreement via Amazon’s cloud provider.
Reports of Anthropic’s AI equipment use in support of US attacks on Iran were also reported by major American media outlets, the Wall Street Journal (WSJ) and Axios.
According to the WSJ, US military command used the AI equipment for intelligence purposes, as well as to assist in target selection and conduct battlefield simulations.
The Claude AI device was also said to be used by US military in operations to apprehend Maduro on 3 January. This reportedly sparked intense disputes between Anthropic and the Trump administration.
Anthropic was said to object, citing its usage terms which do not permit Claude to be used for violence, weapons development, or surveillance purposes.
Since severing ties with Anthropic, other AI companies such as OpenAI have taken over that role. OpenAI’s CEO Sam Altman stated the company had reached an agreement with the Pentagon for use of its equipment, including ChatGPT, within their classified networks.