API
< Back to all clusters
[INTERNATIONAL] · United States, China, Iran · 11 sources

started · updated

US military nearly intercepts Chinese ship due to AI error

An artificial intelligence hallucination nearly triggered a military confrontation between the United States and China. During a period of conflict between the U.S. and Iran, an analyst at the U.S. Special Operations Command Pacific in Hawaii used an AI chatbot to analyze a Chinese vessel's cargo manifest.

The chatbot incorrectly concluded that the ship was transporting components for a nuclear weapons program by combining open-source data with classified signals intelligence. This erroneous finding was then formatted by the AI into an official-looking intelligence report, which was disseminated through military channels.

In response to the report, the U.S. military prepared an interception operation. Armed personnel were ready to board the vessel, and military aircraft had already been deployed into the air. The operation was only halted at the last moment when officials conducted a more thorough review and discovered the intelligence was “entirely false.”

The incident highlights the risks associated with the rapid integration of generative AI into military decision-making processes, specifically the danger of human operators treating fluid, AI-generated text as verified evidence.

Entities

CNN · China · Iran · Pentagon · U.S. Special Operations Command · U.S. Special Operations Command Pacific · United States · United States Special Operations Command · United States military

Claims

What the coverage asserts, and how many sources carry each claim.