A planned US military operation to intercept a Chinese vessel in the Middle East this spring was aborted after intelligence used to justify the mission was found to be based on erroneous artificial intelligence data, according to reports. The vessel was suspected of carrying components linked to a nuclear weapons program, prompting special forces and supporting aircraft to prepare for boarding.

The mission relied on intelligence originating from the US Special Operations Command Pacific in Hawaii, which included details about the ship’s manifest. However, a last-minute review revealed that the information had been generated by an AI chatbot, which produced inaccurate data by conflating or fabricating cargo details. The initial intelligence analyst had inputted raw data into the chatbot, which then created a so-called “hallucination”—a mistaken identification of the vessel’s contents.

Officials halted the operation before it was carried out, recognizing that the intelligence was flawed. One source described the false report as “entirely false” and warned that the misguided action nearly escalated into an armed confrontation between the United States and China. The incident highlights the risks involved in utilizing AI-generated intelligence without sufficient human oversight, particularly amid heightened tensions in the region related to the ongoing conflict with Iran.