The US military almost launched an operation against a Chinese ship due to an AI error

Because of a bug crept into an artificial intelligence intelligence intelligence report, the U.S. military was preparing to intercept a Chinese vessel. This would increase the risk of armed conflict between Washington and Beijing, according to a Minsk-Novosti correspondent.

Photo by pixabay.com

Complicating the already difficult international situation could be the inaccuracy, which consisted in the incorrect definition of cargo carried by a Chinese vessel in the Middle East. According to Bild, a chatbot based on AI technologies matched public information with the intelligence uploaded to it and made a false conclusion, suggesting that components for the nuclear program could be on board. The analyst then once again used AI to form a standard report. So incorrect information in the form of a document that is usually trusted by the military, got to the people who make decisions.

CNN, citing several sources familiar with the situation, reports that an operation was planned to throw armed American soldiers on the ship. However, a re-examination of the data revealed a critical error and avoided undesirable consequences, such as war.

The incident points to the shortcomings of the strategy of accelerated implementation of AI in the activities of the military department, which its head Pete Hegseth presented in early 2026. One of the sensitive areas where artificial intelligence technologies are increasingly in demand is target detection. The problem is not even in the errors of “smart” algorithms, but in the absence of uniform rules for verifying materials from AI to exclude the possibility of civilian casualties or fire on their own. The source of CNN stated that in this area, AI is only shortening the path to bad ideas.