How an AI chatbot’s false nuclear warning almost triggered US-China war in West Asia

In spring, during the conflict with Iran, an intelligence report spread across the US military claiming a Chinese ship in West Asia was transporting components of a nuclear weapons program. This report immediately triggered alarm bells, as military personnel prepared to intercept the vessel, with aircraft already in the air. However, just before the operation commenced, officials reviewed the intelligence and discovered it had been generated by an AI chatbot, which had misidentified the cargo.

The incident almost sparked a war between the US and China. A Special Operations Command analyst fed information about the ship’s manifest into the chatbot, which then produced the false report. The analyst used AI again to turn the findings into an intelligence report, disseminated through the military. Neither Special Operations Command Pacific nor the Pentagon commented when asked by CNN.

The military's reliance on AI stems from its potential to expedite battlefield decisions, from intelligence analysis to routine tasks. Officials also view AI as crucial for maintaining a technological edge over countries like China. In January, Defence Secretary Pete Hegseth unveiled the Pentagon's plan to accelerate AI usage, aiming to deploy advanced AI tools directly to the department's three million personnel.

However, there is no unified set of guidelines governing AI usage in the military, leading to potential errors. This incident highlights the danger of making critical decisions based on misleading AI-generated information.