{"slug": "overreliance-on-palantir-ai-system-contributed-to-us-strike-that-killed-123-says", "title": "Overreliance on Palantir AI System Contributed to US Strike That Killed 123 Iranian Children, Report Says", "summary": "A Pentagon investigation found that US Central Command personnel over-relied on Palantir Technologies' Maven Smart System during the targeting process that led to two Tomahawk missiles striking Shajarah Tayyebeh Elementary School in Minab, Iran, on 28 February, killing at least 123 children and more than 150 people total, according to Bloomberg. Investigators cited years-old intelligence that still classified the site as an Islamic Revolutionary Guard Corps naval facility, a review process compressed from hours to minutes as CENTCOM sought to strike more than 1,000 targets in the opening 24 hours, and civilian-harm mitigation teams cut by Defense Secretary Pete Hegseth from 10 personnel to one, with no team member reviewing the Minab target. Palantir disputed fault, saying it 'is not responsible for the underlying data nor identifying intelligence deficiencies,' and added capabilities after the strike to re-examine underlying intelligence and flag disqualifying information.", "body_md": "# Overreliance on Palantir AI System Contributed to US Strike That Killed 123 Iranian Children, Report Says\n\n## A Pentagon investigation reportedly found that outdated intelligence, compressed targeting reviews and weakened civilian-protection safeguards contributed to the Minab attack\n\nOfficials involved in a Pentagon investigation have revealed that some US Central Command personnel relied too heavily on a Palantir-built AI-enabled system during the targeting process that culminated in a [strike on an Iranian elementary school](https://www.ibtimes.co.uk/pentagon-review-ai-failures-iran-school-strike-1820787), contributing to a chain of failures that killed at least 123 children.\n\nTwo Tomahawk missiles hit Shajarah Tayyebeh Elementary School in Minab on 28 February, killing more than 150 people as the United States opened its military campaign against Iran.\n\nPentagon personnel knew within hours that the US was responsible, *Bloomberg* reported, adding that the Pentagon has yet to release its full investigation or publicly accept responsibility for the strike.\n\n## Personnel Expected AI To Catch Bad Intelligence\n\nAt the centre of the targeting process was Maven Smart System, Palantir Technologies' AI-enabled military platform.\n\nMaven fuses more than 150 different data inputs and allows military personnel to coordinate complex operations ranging from targeting to command and control.\n\nAccording to *Bloomberg*, some US Central Command personnel expected Maven to flag outdated intelligence or inconsistencies in information attached to potential targets.\n\nOfficials involved in the investigation said personnel relied too heavily on that capability, although it remains unclear why they believed the software would perform such checks automatically.\n\nPalantir disputed any suggestion that its software was at fault in the Minab strike. A company spokesperson said Palantir 'is not responsible for the underlying data nor identifying intelligence deficiencies' and said there was no evidence its software was at fault.\n\nPeople familiar with Palantir's Pentagon contracts likewise said the US government retains responsibility for the quality of information fed into Maven.\n\n## School Was Still Listed as Military Site\n\nThe deeper problem began with years-old intelligence. The Minab location had once formed part of an Islamic Revolutionary Guard Corps naval facility, but the site had visibly changed years before the attack.\n\nSatellite imagery showed walls and separate entrances around the school by 2017. By 2018, images showed brightly painted walls, a football pitch and other features indicating the compound's civilian use.\n\nOne US analyst noticed changes as early as 2019, Bloomberg previously reported, but recorded them in a system that was not connected to the primary military intelligence database used for targeting.\n\nAs a result, the Minab site was fed into Maven under its outdated classification as an IRGC facility and ultimately emerged as a recommended target for the first day of the campaign.\n\n## Targeting Work Compressed From Hours to Minutes\n\nInvestigators also found that extraordinary pressure to strike more than 1,000 targets during the opening 24 hours compressed the military's normal review process.\n\nWork that had taken hours in previous conflicts was completed in minutes with Maven, according to people involved in the Pentagon review.\n\nCivilian-protection safeguards had also been weakened. Defence Secretary Pete Hegseth had sharply reduced civilian-harm mitigation teams, with CENTCOM's unit reportedly falling from 10 personnel to one. No member of that team reviewed the Minab target before the strike.\n\n## Palantir Added New Review Capabilities After Strike\n\nFollowing the attack, Palantir added capabilities designed to re-examine underlying intelligence and flag information that could disqualify a target or expose inconsistencies missed during human review, according to a person familiar with the changes.\n\nThe findings have gained further significance after UN investigators said there were reasonable grounds to believe the US strike on the Minab school amounted to the [war crime](https://www.ibtimes.co.uk/trump-weighs-renewed-iran-strikes-un-war-crimes-1820653) of making a civilian object the object of an attack.\n\nThe UN mission said credible sources put the death toll at 157, including 123 children. The Pentagon says its investigation remains ongoing.\n\nWhat has emerged so far points not to one malfunctioning algorithm, but to a broader failure in which stale intelligence, rapid decision-making, diminished civilian-protection resources and misplaced expectations about what AI could catch all converged before the missiles were launched.\n\n© Copyright IBTimes 2026. All rights reserved.", "url": "https://wpnews.pro/news/overreliance-on-palantir-ai-system-contributed-to-us-strike-that-killed-123-says", "canonical_source": "https://www.ibtimes.co.uk/palantir-ai-system-us-strike-iranian-school-1821738", "published_at": "2026-09-24 14:13:57+00:00", "updated_at": "2026-09-24 14:29:45.911603+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-safety", "ai-policy", "ai-ethics"], "entities": ["Palantir Technologies", "Maven Smart System", "US Central Command", "Pentagon", "Shajarah Tayyebeh Elementary School", "Minab", "Pete Hegseth", "Bloomberg"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/overreliance-on-palantir-ai-system-contributed-to-us-strike-that-killed-123-says", "markdown": "https://wpnews.pro/news/overreliance-on-palantir-ai-system-contributed-to-us-strike-that-killed-123-says.md", "text": "https://wpnews.pro/news/overreliance-on-palantir-ai-system-contributed-to-us-strike-that-killed-123-says.txt", "jsonld": "https://wpnews.pro/news/overreliance-on-palantir-ai-system-contributed-to-us-strike-that-killed-123-says.jsonld"}}