A Chatbot-Assisted Intelligence Report Sent Armed U.S. Troops to Board a Chinese Ship for Nuclear Weapons Parts That Were Never There — and the Top Democrats on Intelligence, Armed Services and Defense Appropriations Demand an Inspector General Investigation of AI Errors in Military Targeting
On September 18, 2026, CNN reported that earlier this year, during the Iran war, a U.S. Special Operations Command analyst used an AI chatbot to help produce an intelligence report claiming a Chinese ship in the Middle East was carrying components for a nuclear weapons program. The report circulated across the military; armed personnel suited up to board the vessel and military aircraft were already airborne before senior officials examined the underlying sourcing and concluded the report was "entirely false." One source told CNN the "hallucination" reflected a trend rather than an isolated incident, and another said it "almost started a war." On September 19, Senate Intelligence Committee Vice Chairman Mark Warner, Armed Services Ranking Member Jack Reed and Defense Appropriations Subcommittee Ranking Member Chris Coons wrote to Defense Secretary Pete Hegseth and Director of National Intelligence Jay Clayton demanding an "immediate investigation by relevant Inspectors General." The letter also cited the February strike on a school in Minab, Iran, which killed nearly 200 people, after commanders bypassed warnings in AI-powered databases that the target intelligence was badly outdated. The Pentagon, which under Hegseth has pushed to accelerate AI adoption across operations, had not publicly responded as of September 19.
“[We call for an] immediate investigation by relevant Inspectors General, with a higher degree of public transparency in the ultimate findings, given the impacts these perceived failures have on public confidence in our intelligence and warfighting missions.”