AI hallucination nearly triggered US military strike on Chinese ship
Original: AI hallucination nearly triggers US military operation
Why This Matters
Shows that AI hallucinations in high-stakes military contexts can trigger real-world escalation before errors are caught.
In spring 2026, U.S. military aircraft were airborne for an armed operation against a Chinese vessel before officials discovered the intelligence was hallucinated by an AI chatbot. The operation was aborted at the last minute, averting a potential conflict with China, CNN reported on September 18.
During the U.S. war with Iran, a Special Operations Command analyst used an AI chatbot to synthesize open-source data with classified signals intelligence. The chatbot misidentified a Chinese ship's cargo manifest, falsely indicating it carried components for a nuclear weapons program. The analyst then used the same tool a second time to format the erroneous findings into an official-looking intelligence summary, which circulated through command channels. By the time officials identified the error, military aircraft were already in the air. The operation was called off just in time.
The incident highlights a structural risk in AI-assisted military decision-making: hallucinated outputs can move up the chain of command faster than human review catches them. The Pentagon has publicly championed AI as a way to accelerate its 'kill chain,' but this case shows that same speed can bypass verification steps.
Jake Steckler, a research scholar at GovAI and former U.S. Army officer, told TechCrunch that service members need to understand the uncertainty inherent to large language models—especially in targeting, intelligence analysis, and operational planning. He argued the incident should prompt stronger safeguards, not a pullback from AI altogether. 'Prioritizing adoption speed over all else will likely lead to incidents that only make service members lose trust in these systems,' he said.