AI nuclear error 'almost started WAR with China' as US forces scrambled to intercept ship

The extraordinary incident unfolded earlier this year after a US special operations analyst used an AI chatbot to analyse information relating to the ship's cargo.

The technology wrongly concluded that the vessel, which was travelling through Middle Eastern waters, was transporting components linked to a nuclear weapons programme.

Rather than ending there, AI was reportedly used again to turn the initial finding into a formal intelligence assessment.

The report was then circulated across the US military, prompting officials to begin planning an operation to intercept the vessel.

Aircraft were dispatched and troops began preparing to board the ship.

But shortly before the operation was due to begin, officials subjected the intelligence assessment to further scrutiny.

AI wrongly concluded that the vessel was transporting components linked to a nuclear weapons programme (file image)

GETTY

They discovered that AI had been used to produce the report and subsequently determined that the technology had misidentified the material aboard the vessel.

One source familiar with the incident described the AI-generated intelligence to CNN as "entirely false".

The source added that it "almost started a war" with China, given the potential consequences of US forces taking military action against a Chinese vessel.

The revelations come just days before Chinese President Xi Jinping is due to travel to Washington for talks with President Donald Trump.

The revelations come just days before Xi Jinping is due to travel to Washington for talks with President Donald Trump

GETTY

Artificial intelligence, trade and critical minerals are expected to feature in discussions between the two leaders.

The near-miss has also thrown a spotlight on the rapidly expanding use of AI across the US military.

Officials have deployed the technology to analyse vast quantities of intelligence, assist with identifying potential targets and streamline military logistics.

One source told CNN: "AI in targeting is definitely something that is ramping up and there is no real guidance for how having a human in the loop will prevent civilian casualties or fratricide."

It remains unclear whether the chatbot responsible for the false assessment was an internal Pentagon system or commercially available technology.

One former US official told the outlet: "The internal tools are mostly just copies of the commercial stuff wearing lipstick."

Secretary of War Pete Hegseth has championed the rapid adoption of artificial intelligence throughout America's armed forces.

Mr Hegseth said in January: "We will unleash experimentation, eliminate bureaucratic barriers, focus our investments and demonstrate the execution approach needed to ensure we lead in military AI."