An AI Error Nearly Led To A War Between The U.S. And China In The Spring Of 2026
- 19.09.2026, 9:57
A Pentagon analyst placed too much trust in the conclusions of a neural network.
In the spring of 2026, the U.S. military nearly launched an operation against a Chinese vessel in the Middle East after receiving an intelligence report containing false information generated by artificial intelligence. This was reported by CNN citing four sources familiar with the situation.
The incident occurred against the backdrop of the war with Iran. A report circulated among U.S. military personnel claimed that the Chinese vessel was transporting nuclear weapons components. The report immediately raised alarm, after which the military began preparing to intercept the ship.
According to CNN sources, armed U.S. service members were preparing to board the ship, and military aircraft were already in the air. The situation changed only shortly before the operation was set to begin, when officials more thoroughly verified the source of the information.
It turned out that the report had been prepared by an analyst with the Special Operations Command using an AI chatbot. It was the AI that incorrectly identified the nature of the cargo on the Chinese vessel. CNN was unable to determine exactly which cargo had been misidentified.
One source called the document “completely false,” but the consequences could have been far more serious. According to the source, the error “nearly led to war.” Any U.S. operation against the Chinese vessel could have led to further escalation between Washington and Beijing.
According to CNN, the analyst first asked a chatbot to analyze intelligence regarding the ship’s cargo manifest, which had been provided by U.S. Pacific Command’s Special Operations Command. The chatbot cross-referenced open-source information with intelligence data obtained through electronic intelligence, after which it reached an erroneous conclusion about the cargo.
The analyst then used the AI again—this time to format the findings into a standard intelligence report. The document was circulated among military officials, who could have interpreted it as standard, verified analytical material.
The incident occurred against the backdrop of the widespread adoption of AI within the U.S. armed forces. The technology is intended to be used for analyzing large volumes of intelligence, selecting targets for strikes, logistics, and other tasks. In January, U.S. Secretary of Defense Pete Hagset presented a strategy for the accelerated adoption of artificial intelligence.
At the same time, according to CNN, the use of AI in the military and intelligence sectors remains decentralized: different units use different systems and follow different security protocols. There is no uniform standard for verifying information generated by such tools.
According to CNN sources, instances of so-called AI “hallucinations”—where the system generates false information—have already occurred within the U.S. intelligence community. Some officials fear that the push to speed up intelligence processing could lead to more errors.
Advertisement
“AI allows you to arrive at a bad idea faster,” said one of the publication’s sources.