By Anjali Sharma
WASHINGTON – According to AI-assisted intelligence report on Saturday said that it falsely identifying a Chinese vessel in West Asia as carrying components linked to a nuclear weapons programme reportedly brought the US military close to launched an operation against the ship during the war with Iran.
The analyst reportedly used artificial intelligence twice first to assess intelligence about the vessel’s manifest and later to package the findings into a formal report that circulated through military channels.
Armed US personnel were preparing to board the vessel and military aircraft had already taken to the air when officials scrutinised the intelligence and discovered a serious problem: the assessment had been produced with the help of an AI chatbot that had misidentified the ship’s cargo, media reported, cited multiple sources familiar with the episode.
One source familiar with the incident described the intelligence as “entirely false” and said the episode “almost started a war”.
Media reports said it could not establish what the Chinese vessel was actually carrying.
The intelligence trail began with reporting concerning the vessel’s manifest that originated with US Special Operations Command Pacific in Hawaii, according to media
An analyst queried a chatbot about the intelligence. The AI system reportedly drew on open-source material as well as classified signals intelligence held by the government.
It reached the wrong conclusion about the cargo.
The analyst then turned to AI again to package the findings into a standard intelligence report. That report travelled through military channels.
It remains unclear whether the chatbot was a commercially available system or one developed by the US government.
The report set off alarm within the military because it claimed the Chinese vessel was transporting material associated with a nuclear weapons programme.
Four sources cited by media said the US military began moving towards an operation against the ship.
Two said armed personnel were preparing to board it. Two others said military aircraft were already airborne.
The operation did not go ahead. Officials examined the intelligence more closely shortly before it was due to begin and traced the assessment back to the AI-assisted analysis.
The potential consequences were stark, according to one of media sources.
The source said a US operation against the Chinese vessel could have escalated into an armed confrontation between the two countries.
The episode has emerged as the US military rapidly expands its use of artificial intelligence across intelligence and battlefield functions.
US Secretary of War Pete Hegseth unveiled the Department of War’s Artificial Intelligence Acceleration Strategy.
The plan covers warfighting, intelligence, and enterprise operations and seeks to accelerate the deployment of AI across the department.
“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 while announcing the strategy.
The department has continued expanding the technology’s role. Its Agent Network programme, launched in June, uses AI-enabled tools for battle management, decision support and targeting.
The department said the system does not autonomously select or strike targets and that commanders retain control of decisions.
In September, ChatGPT and Grok were added to GenAI.mil alongside Gemini, expanding the AI systems available within the department.
Media report points to a separate problem: there is no single standard across the government for verifying information produced by the different AI systems being used.
Many officials cited in the report were concerned that pressure to use AI to produce intelligence more quickly could also increase the risk of mistakes entering formal assessments.
The younger analysts were described by sources as being more comfortable with the technology and potentially more inclined to trust its output without adequate scrutiny.
“AI allows you to get to a bad idea faster,” one source told media outlet
Another source raised concerns about the safeguards surrounding AI-assisted targeting.
“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.”
According to media reports one source said the Chinese vessel episode was not necessarily an isolated instance of AI hallucinations entering intelligence work as the technology becomes more widely used across the US government.
