U.S. probe links Minab school strike to outdated targeting data, not AI error
A preliminary U.S. military investigation into last month’s airstrike on an elementary school in Minab, southern Iran, indicates the operation may have relied on outdated targeting data, suggesting human error rather than a technical AI fault. The strike killed around 170 people, according to local reports.
The New York Times, citing current and former U.S. officials, reports that CENTCOM built coordinate targets for a Revolutionary Guard Navy base near the school using data from the Defense Intelligence Agency that had not been updated to reflect changes at the site. The analysis suggests the school stood on or near land that had previously housed a military facility, a fact not reflected in the targeting data.
Satellite imagery analyzed for the investigation shows the school appears to have been established between 2013 and 2016. The school buildings are described as separated from the nearby base by fencing, with three entrances added since then, and watchtowers around the school removed. These changes indicate a mismatch between current ground realities and the data used to plan the strike.
The inquiry remains in a preliminary stage. Investigators are examining why outdated data were used without proper verification, whether updated information existed within the Defense Intelligence Agency, and who within the chain of data validation approved the targeting coordinates. The probe also looks at the role of the National Geospatial-Intelligence Agency in reviewing satellite imagery for the operation.
Officials say technological error involving an artificial intelligence targeting program is being considered, but the current assessment points more toward human error or mis-provision of data rather than a software flaw.
As context, researchers point to a historical example: during the 1999 Kosovo War, the United States bombed the Chinese embassy in Belgrade while attempting to strike a weapons facility. The CIA later acknowledged that staff shortages had led to incomplete data maintenance, contributing to a mis-targeting that killed three people. The Minab incident underscores enduring lessons about accuracy, verification, and civilian risk in wartime targeting.
For U.S. readers, the broader takeaway is clear: civilian casualties from targeted strikes depend on the reliability of intelligence data and verification processes. The episode has potential implications for U.S. defense and intelligence practices, the risk calculus of future operations in the Middle East, and the stability of supply routes and markets tied to the Persian Gulf region. It also highlights how evolving technologies intersect with traditional data-management challenges in high-stakes military decisions.