US nearly raided Chinese ship after AI falsely flagged ‘nuclear’ cargo for Iran — CNN

Report says warplanes were already in the air when officials realized intel report was written by a hallucinating chatbot; it 'almost started a war,' source says

Illustrative: A Chinese container ship is seen at the Panama Canals’ Cocoli locks, in Panama City, December 3, 2018. (AP Photo/Arnulfo Franco)

A US intelligence report produced by AI with false information nearly led to a confrontation between Washington and Beijing after it mistakenly surmised that a Chinese ship was carrying nuclear components to Iran during the Middle East war, CNN reported over the weekend, citing sources familiar with the episode.

The US military prepared to seize the vessel, and warplanes were in the air, CNN said, before officials examined the report more carefully and realized it was written with the help of a chatbot that misidentified the cargo. CNN said it was unable to ascertain what the actual cargo was.

CNN quoted one source as saying the intelligence report was “entirely false” but “almost started a war.”

According to the network, a US military analyst had queried the chatbot about intelligence from the Hawaii-based US Special Operations Command Pacific about the ship, and the chatbot based its faulty response on both open-source and secret signals intelligence.

It was unclear whether the bot was a commercial or government product, CNN says, adding that neither the Pentagon nor Special Operations Command Pacific responded to a request for comment.

US Central Command — which is responsible for American forces in the Middle East — did not immediately respond to a request for comment on the reported incident.

A US Air Force F-16 Fighting Falcon aircraft pulls away from a KC-135 Stratotanker aircraft in the US Central Command area of responsibility May 21, 2026. (US Air Force/ Adriana Jordan-Alcaniz)

The CNN report comes after a furor erupted earlier this month when a researcher publicly resigned from leading AI firm Anthropic over fears that the technology could escape human control.

Another Anthropic employee, who did not resign, said publicly that “we really do earnestly believe AI could kill all humans,” and that he thought the chances were greater than “10 percent within the next decade.”

Despite the warnings, US President Donald Trump and some of his key allies have voiced opposition to drastically curbing the development of artificial intelligence.

In a separate incident involving the US military and China, Politico reported Friday that F-35 jet parts — including a sensitive stealth canopy for a cockpit — were diverted to Hong Kong and are now missing, raising fears that Beijing had obtained sensitive American military technology.

The parts were headed from Australia to the United States for repairs when the diversion occurred, the US outlet reported, adding that it was unclear why the parts were rerouted or where they are now.

A spokesperson for Lockheed Martin, which manages the F-35 supply chain, said the company could not discuss specific shipments because of “security reasons.”

“Lockheed Martin follows all US regulations established by the Department of War and regulated by the Defense Logistics Agency (DLA) regarding the shipment of aircraft components,” the spokesperson said.

“Our teams handle every shipment with the utmost diligence and safeguards to ensure the integrity of the F-35 program and the security of our allied partners.”

read more: