Exclusive to Afaq News: An AI Error Nearly Sparked a War Between America and China.. Here are the details

- An intelligence report, circulated among the U.S. military in the spring of this year, coinciding with the war with Iran, triggered an immediate state of alert, as it indicated that a Chinese ship in the Middle East was transporting components related to a nuclear weapons program. According to four sources familiar with the incident, the U.S. military moved to devise plans to intercept the ship, and armed elements of the military were even preparing to board it.

Two sources, one of whom participated in the previous information, reported that military aircraft were already flying in the air. Officials only discovered the truth just before the planned operation began, as they scrutinized the report prepared by an analyst from the Special Operations Command, which turned out to have been produced with the help of artificial intelligence, and that the 'chatbot' used by the analyst had incorrectly identified the nature of the materials the ship was carrying.

The Afaq News network was unable to determine the nature of the cargo that was misidentified. One source described the report as 'completely wrong,' while also noting that it 'almost ignited a war,' as any U.S. operation against a Chinese ship would have involved the risk of sliding into an armed conflict between the two countries.

U.S. military officials and the intelligence community are seeking to integrate artificial intelligence into nearly every aspect of their work, from analyzing the vast amounts of raw intelligence data collected by the United States and identifying strike targets, to more routine applications such as budget management, logistics, and supply chains. However, this incident highlights the significant risks of using this powerful and modern technology, which is still not well understood, in targeting operations during wartime.

Analysts have long warned that artificial intelligence could lead to catastrophic miscalculations if countries rely on weak or erroneous data; the kind of miscalculation that could prompt the United States to fire on a Chinese ship based on inaccurate information. In this particular incident, the analyst posed a question to a chatbot regarding intelligence information about the ship's cargo list, which had originally been issued by the U.S. Special Operations Command in the Pacific (based in Hawaii).

It is unclear whether the chatbot used is commercially available or affiliated with the U.S. government. A former senior U.S. official familiar with the artificial intelligence systems used by military analysts and intelligence agencies said, "Most internal tools are just copies of commercial software that have been given superficial improvements."

The chatbot integrated intelligence information available from open sources with classified intelligence gathered through communications signals in the government's possession, and reached its fateful conclusion about the nature of the materials the ship was carrying. The analyst then used artificial intelligence again to formulate the findings into a standard intelligence report, the type of report that military officials trust, which was then disseminated.

The U.S. Special Operations Command in the Pacific and the U.S. Department of Defense did not respond to Afaq News' request for comment. The justification for the rush to adopt artificial intelligence lies in its ability to assist the military in making field decisions, such as identifying targets to strike or determining the locations of military asset transfers, at a faster pace.

Officials say the United States cannot afford to delay the integration of artificial intelligence, anticipating that it may one day have to engage in a war against China or another adversary that could potentially stay one step ahead of the United States. In January, U.S. Defense Secretary Pete Hegseth issued a strategy for his department titled "Artificial Intelligence Acceleration Strategy," in an effort to accelerate the military's use of this technology.

Hegseth said in a speech announcing the strategy: "We will unleash experimentation, remove bureaucratic obstacles, focus our investments, and demonstrate the necessary implementation approach to ensure our leadership in military artificial intelligence." The strategy also pushes for the widespread use of artificial intelligence throughout the military, as the department orders the availability of artificial intelligence across several programs, aiming to "make artificial intelligence experiences and digital transformation accessible to everyone within the department, by placing world-leading U.S. artificial intelligence models directly in the hands of our three million civilian and military personnel, across various security clearance levels," according to the memorandum announcing the strategy.

However, these efforts are characterized by decentralization - according to U.S. officials familiar with the operations - as various government entities use diverse tools and are subject to varying orders and safety standards. There is no unified set of standards that the United States follows to verify the accuracy of the information generated by these tools; the multitude and diversity of artificial intelligence systems deployed by different sectors within the military and intelligence agencies mean that levels of reliability and effectiveness vary significantly.

For weeks, policymakers in Washington have engaged in intense discussions about artificial intelligence, following a series of serious warnings issued by engineers and executives in the technology sector in Silicon Valley regarding the potential for artificial intelligence to spiral out of human control, and the consequences that could threaten civilization. However, the incident involving the Chinese ship highlights a different and more urgent danger: that humans may make catastrophic decisions based on inaccurate or misleading information produced by artificial intelligence or other automated systems.

Sources said the military is rapidly moving towards using artificial intelligence to assist in target identification, a field that carries clear risks of fatal errors. Another source familiar with current military policies stated: "The use of artificial intelligence in target identification is expanding, and there are no actual guidelines outlining how the presence of a human element in the decision-making loop can prevent civilian casualties or friendly fire incidents."

According to one source, the type of "hallucination" (generating false information) produced by the tool used by the analyst in this case was not an isolated incident within the intelligence community since the onset of these tools in government institutions. For some older intelligence officials, even those who broadly support the use of artificial intelligence within the military, AI has pressured analysts to produce and disseminate intelligence information at a faster pace, opening the door to errors.

Several sources pointed out that younger analysts, in particular, have grown up accustomed to using these tools, making them more inclined to trust them without subjecting them to critical evaluation. One source said: "Artificial intelligence allows you to access a bad idea more quickly."