Pentagon Used AI Model ‘Claude’ in Attempted Maduro Kidnapping, Reports Claim

The Pentagon reportedly utilized the AI model Claude during a January 3 mission in Venezuela aimed at capturing President Nicolás Maduro, sources cited by The Wall Street Journal and Axios claim. According to the reports, the AI was employed not only during planning but also throughout the operational phase, although the precise details of its use remain undisclosed.

Claude, developed by the U.S.-based AI company Anthropic, is governed by strict internal policies that prohibit its deployment for violence, weapon production, or surveillance purposes. Despite this, Pentagon officials allegedly integrated the AI into their classified systems for the first time, facilitated through Anthropic and another technology provider, Palantir.

While no U.S. personnel were reportedly killed during the operation, multiple Venezuelan and Cuban security forces allegedly lost their lives, according to several media sources.

Anthropic has declined to comment on the specifics of Claude’s use in the mission, emphasizing that any deployment must adhere to company policies.

Venezuelan Acting President Delcy Rodríguez, in an interview with the BBC, confirmed that Maduro remains the legitimate head of state. She emphasized that his temporary absence has not altered the constitutional functioning of the government. Rodríguez also stated that post-operation discussions with the United States involved agreements on oil and other political matters, while asserting that both President Maduro and First Lady Cilia Flores remain innocent.

U.S. President Donald Trump claimed on Friday that relations with Venezuela are improving and highlighted U.S. assistance to the country across oil and other sectors, mentioning potential plans for a visit. The U.S. State Department also announced that more than six tons of priority medical supplies had been sent to Venezuela in 25 pallets as part of stabilization efforts, one month after the special forces operation.