Pentagon investigation links AI reliance to fatal Iran school strike

A Pentagon investigation has linked a February 28 missile strike on the Shajarah Tayyebeh elementary school in Minab, Iran, to a heavy reliance on artificial intelligence and outdated intelligence. The strike, involving two Tomahawk missiles, killed more than 150 people, including at least 123 children.

The inquiry found that while the location had been classified as a military facility years prior, commercial imagery and analyst observations as early as 2019 indicated its transition to a school. However, these updates were not integrated into the primary military intelligence databases used for targeting. The compressed pace of the air campaign, which hit over 1,000 targets in its first 24 hours, reduced the time available for final verification checks.

The Maven Smart System, an AI-enabled platform by Palantir, was used to process data during the targeting process. While Palantir stated its software was not at fault for intelligence gaps, the incident has prompted U.S. Central Command and the Pentagon to overhaul targeting procedures. New upgrades to the Maven system aim to allow the AI to flag anomalies and inconsistencies in intelligence, while the military has increased the use of open-source intelligence to better identify civilian activity around potential targets. Pentagon officials emphasized that human commanders retain final decision-making authority for all strikes.