The United States military used Elon Musk’s artificial intelligence technology ‘Grok’ to conduct strikes on more than 2,000 Iranian targets within four days, according to a written testimony by Cameron Stanley, head of Digital and Artificial Intelligence at the Pentagon. The disclosure came during a civil case in the Northern District Court of Mississippi, where the Trump administration intervened in an environmental lawsuit filed against Musk’s company xAI.

Stanley stated that the US government deployed a ‘Grok Gov Model’ based on xAI’s commercial version, integrated into the Pentagon’s Maven Smart Systems (MSS) for target identification, intelligence gathering, and military readiness. The testimony did not specify the exact dates of use, leaving unclear whether the AI was linked to a February 28 strike that killed 156 civilians, including 120 schoolchildren. Iran’s Foundation of Martyrs and Veterans reported about 3,500 civilian deaths since the joint US-Israeli offensive began.

The lawsuit, filed by the NAACP on April 26, accuses xAI of violating environmental laws by operating 27 methane gas turbines in Southaven, Mississippi, to power the ‘Colossus 2’ data center. The Pentagon and the US Department of Justice opposed the case, citing national security risks if the data center’s operations were disrupted.