US military AI error nearly sparked conflict with China over false nuclear cargo report

A US military artificial intelligence error nearly triggered an armed conflict with China in spring 2026, according to CNN. A chatbot falsely identified a Chinese freighter's cargo as parts for a nuclear weapons program, prompting armed soldiers to prepare for a boarding operation. The incident occurred during the war with Iran when an intelligence report claimed the ship in the Middle East carried nuclear components. US military aircraft were already airborne before officials discovered during a final review that the chatbot had misidentified the cargo, though the actual contents remain undisclosed. An analyst at US Special Operations Command Pacific had used the AI to evaluate the ship's manifest, combining public data with classified surveillance information. The system hallucinated the threat, and the analyst later used AI to format the false findings into an official report. US intelligence sources say such AI hallucinations are not isolated, with younger analysts particularly prone to uncritical trust in the technology.