U.S. Nearly Boarded Chinese Ship In Middle East After AI-Assisted Intelligence Error

The U.S. military came close to intercepting and boarding a Chinese vessel in the Middle East this spring after an intelligence report wrongly said the ship was carrying components linked to a nuclear weapons programme, CNN reported.

The intelligence came from the U.S. Special Operations Command Pacific in Hawaii, CNN reported.

An analyst used a chatbot to examine information about the Chinese vessel’s manifest. The material included publicly available information, classified signals intelligence and other data held by the U.S. government.

The AI system wrongly identified the cargo as components linked to a nuclear weapons programme. The analyst then used the system to help prepare the findings as a standard intelligence report.

The report was circulated through military channels but was later found to be wrong.

One source described the assessment as “entirely false” and said it had “almost started a war”, according to CNN.

The intelligence report led the U.S. military to prepare an operation to stop the vessel.

According to CNN, U.S. military aircraft were already airborne while armed personnel prepared to board the Chinese ship.

The operation was stopped after more experienced analysts and subject-matter experts reviewed the information behind the report and found that the AI-generated identification of the cargo was wrong.

CNN reported that the error was discovered as the operation was about to begin. The identity of the vessel and its actual cargo have not been disclosed.

A U.S. attempt to stop and board a Chinese commercial vessel could have created a direct confrontation between Washington and Beijing. The incident took place during the ongoing conflict with Iran.

The operation, however, was called off after the intelligence was found to be inaccurate.

US military expands AI use

The Pentagon has been using AI across military operations. Its January 2026 AI strategy calls for faster integration of AI capabilities across warfighting, intelligence and other missions, including the GenAI.mil initiative.

The U.S. military has said AI can help process large amounts of information and support faster decision-making.

In January, Defence Secretary Pete Hegseth announced an “Artificial Intelligence Acceleration Strategy” aimed at expanding AI use across the U.S. Department of Defense and reducing bureaucratic barriers.

GenAI.mil is intended to give military and civilian personnel access to commercial AI models. An accompanying memo called for AI models to be made available across the department at different classification levels.

Concerns over AI-generated intelligence

Sources cited by CNN said AI use across the U.S. government remains decentralised, with different organisations using different systems and following different safety rules and restrictions.

Officials have also raised concerns about the lack of a common standard for checking AI-generated information before it is distributed.

Some sources told CNN that younger analysts may be more likely to accept AI-generated information without enough scrutiny.

Other intelligence officials have warned that pressure to produce information faster could result in assessments being distributed before they have been properly checked.

A former senior U.S. official familiar with military AI systems told CNN that some internal tools are closely based on commercial AI technology.

The former official said these systems can combine open-source information with classified signals intelligence. That combination, the official said, can make incorrect conclusions appear more reliable.

References: India Today, The Independent