AI nearly triggered US-China conflict over fake nuclear weapons
US intelligence nearly seized a Chinese vessel (photo: Pexels)
The US Navy nearly intercepted a Chinese civilian vessel in the Middle East because of a fabricated AI-generated report. A chatbot hallucination about nuclear components being on board nearly triggered a major military conflict, according to CNN.
How did an AI error nearly trigger a military assault?
US military intelligence prepared a classified analytical report claiming that a Chinese vessel in the Middle East was allegedly transporting components for building nuclear weapons.
The Pentagon immediately developed an operation to detain the vessel: combat aircraft were scrambled, while boarding teams prepared to go aboard.
The mission was canceled at the last moment when officials decided to recheck the sources. An additional review showed that the information about the nuclear cargo had been completely fabricated by AI algorithms.
According to sources in the Navy, seizing the ship by force could have triggered a direct military confrontation between nuclear-armed states.
Why do chatbots make critical intelligence errors?
The incident was caused by the widespread deployment of AI tools at all levels of analysis without an adequate verification system.
It became known that one employee used a chatbot to combine open-source data with classified signals intelligence. The algorithm generated a fabricated conclusion and presented it as fact.
Experts identify several systemic problems with the military’s use of AI:
Decentralization of tools — different units use different programs without a unified data verification standard.
Use of commercial solutions — intelligence agencies often use adapted versions of ordinary civilian chatbots.
Persistent hallucinations — the incident involving the Chinese vessel turned out to be just one in a series of false conclusions generated by AI when analyzing large volumes of information.
Experts emphasize that as the volume of data grows, the military is increasingly relying on AI algorithms to identify targets. However, the lack of strict oversight creates a risk of catastrophic errors in the future.
You may also be interested in:
