The US 'almost started a war' with China after artificial intelligence incorrectly analyzed that the contents on a Chinese ship included nuclear weapon components.

This spring, a US special operations analyst used an AI chatbot to analyze the Chinese ship's manifest, multiple sources told CNN. 

It incorrectly concluded that the vessel, then transiting the waters in the Middle East, was shipping components of a nuclear weapons program.

Later, the analyst again used AI to develop that determination into an intelligence report that was later circulated across the military. 

Almost immediately, military planners began drafting plans to intercept the Chinese ship. The plan was to board the ship, and planes were in the air for the mission as soldiers were preparing. 

However, just before the mission was set to launch, officials further reviewed the intelligence assessment and found it was created with the help of AI. They soon realized that the tech had misidentified the material the ship was carrying. 

One source called the AI report 'entirely false.' It also, they added, 'almost started a war' with China, as any military action against the ship could have sparked a conflict between the two superpowers. 

The revelations of the nearly calamitous operation come less than a week before Chinese President Xi Jinping is set to visit Washington to meet with President Donald Trump to discuss AI, trade, critical minerals and more. 

The Trump administration has rolled out AI use across the federal government, and the Pentagon has been a voracious consumer of the new tools. 

Officials have used AI to analyze large amounts of raw intelligence and determine which targets to strike. It has also been used to streamline logistics and help balance budgets. 

'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,' one source revealed. 

The scary incident underscores the rapid deployment of AI across the military and the risks it poses in critical intelligence work. 

It is unclear whether the AI used was an internal Pentagon system or a commercial one. 

'The internal tools are mostly just copies of the commercial stuff wearing lipstick,' one former US official told CNN. 

Secretary of War Pete Hegseth has advocated for the rapid adoption of AI among the armed forces, an initiative that went into overdrive earlier this year.  

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

Recently, the White House has been grappling with how to secure American AI dominance while protecting citizens against chatbots running amok. 

OpenAI and Anthropic, the companies behind ChatGPT and Claude, have warned in recent weeks that AI firms may benefit from a slowdown in AI development, citing safety risks. 

The latest incident with the Chinese vessel highlights exactly how AI can create inaccurate or misleading information that soldiers or sailors may then act on.