AI-generated intelligence error nearly triggers U.S. military operation against Chinese ship

The United States military nearly launched a mission to intercept a Chinese vessel in the Middle East due to a false intelligence report generated by artificial intelligence. The incident occurred during a period of conflict involving Iran, a time when preventing Tehran from acquiring nuclear weapons was a primary U.S. objective.

A U.S. intelligence analyst used an AI chatbot to analyze data collected by the U.S. Special Operations Command's Pacific unit, specifically regarding a Chinese ship's cargo manifest. The chatbot incorrectly concluded that the vessel was transporting components essential for nuclear weapons production. This erroneous finding was compiled into a formal intelligence report and distributed through classified Pentagon channels.

As the report was treated as a reliable human-authored document, military preparations for a boarding operation commenced. Special forces were reportedly ready to board the ship, and aircraft were deployed for support. The operation was halted only moments before execution when officials conducted a deeper review and discovered the report was based on AI-generated hallucinations. An anonymous official stated the intelligence was “completely false” and “nearly started a war.”