The US military nearly launched an operation against a Chinese vessel this spring based on an intelligence report generated with the help of artificial intelligence that turned out to be "entirely false," according to a CNN investigation published Friday, September 18, citing four sources familiar with the incident.
What the False Report Claimed
The AI-generated intelligence report indicated that a Chinese vessel in the Middle East was carrying components of a nuclear weapons program. The claim surfaced amid the ongoing tensions tied to the US-Iran war, and the report was widely disseminated within the military as a credible, standard intelligence assessment.
How Close the Military Came to Acting
Based on the report, the US military scrambled jets and armed service members prepared to board the vessel. Military planes were already in the air and personnel were ready to move on the ship. It was only just before the planned operation that officials took a closer look at the underlying report and discovered it had been generated with AI assistance — and that the chatbot had inaccurately identified the ship's actual cargo. CNN was unable to determine what the vessel was actually carrying.
The Stakes If It Had Gone Ahead
One source told CNN the report was "entirely false" but also "almost started a war" — a stark characterization given that any US operation against a Chinese vessel could have risked spiraling into a direct armed confrontation between the two nations, both of which are nuclear powers, at a moment when the US was already engaged in an active conflict elsewhere in the region.
How the Error Happened
According to CNN's sources, a special operations command analyst asked an AI chatbot to analyze intelligence reporting related to the Chinese ship's manifest. It remains unclear whether the chatbot was a commercially available product or a government-built tool. The chatbot combined open-source information with secret signals intelligence and arrived at an incorrect conclusion about the vessel's contents. The analyst then used AI again to package those findings into a standard, formally formatted intelligence report — the kind of document military officials are trained to trust — and disseminated it through normal channels.
A Blunt Assessment of Internal Military AI Tools
A former senior US official told The Daily Beast that "the internal tools are mostly just copies of the commercial stuff wearing lipstick," a pointed critique suggesting the military's own AI systems may not offer meaningfully more rigor or reliability than off-the-shelf commercial chatbots, despite being deployed in far higher-stakes contexts.
The Broader Push Behind the Incident
The near-miss occurred amid a much broader effort across the US military and intelligence community to weave AI into nearly every facet of operations — from analyzing massive volumes of raw intelligence and selecting strike targets, to more mundane tasks like budgeting, logistics, and supply chain management. In April, a senior Pentagon official reportedly told DefenseScoop that Defense Secretary Pete Hegseth's department was "cattle-driving, as we say, everything in GenAI.mil" — language reflecting an aggressive, rapid-adoption posture toward AI integration across the department.
Why the Military Is Moving So Fast
Officials have justified this rapid rollout on competitive grounds: the reasoning is that AI can help the military make battlefield decisions — including which targets to strike or which assets to move where — significantly faster than human analysis alone, and that the US can't afford to fall behind in AI integration in case it must one day confront China or another adversary capable of moving faster using similar tools.
The Pentagon's Response
An official with the Department of War (the Pentagon's rebranded name under the current administration) told TMZ: "On any given day, the U.S. military monitors, assesses, and analyzes hundreds of potential threats around the world to ensure the defense of our nation and our allies. Out of operational security and as a matter of policy, we do not comment on specific intelligence reports, alleged operational planning, or ongoing military assessments." The statement continued: "The Department of War remains committed to the rigorous verification of all intelligence inputs and the responsible integration of emerging technologies to support our service members in making informed, precise, and safe decisions in any operational environment." US Special Operations Command Pacific and the Pentagon did not respond directly to CNN's request for comment.
Timing Within a Broader AI Safety Debate
CNN noted the revelation lands amid weeks of intense Washington policy debate over AI, fueled by a series of warnings from Silicon Valley engineers and tech CEOs about the risks of insufficiently verified or safeguarded AI systems being deployed in high-stakes contexts — a debate this specific incident is likely to intensify, given it represents a concrete, real-world case of AI-generated misinformation nearly triggering a military confrontation between major powers.
What's Still Unknown
Key details remain unconfirmed, including which specific AI tool was used, what the ship was actually carrying, and whether any disciplinary or procedural changes have been implemented since the incident to prevent similar errors. Neither the Pentagon nor US Special Operations Command Pacific has provided additional detail beyond the general statement given to TMZ.
What's Next
With this incident now publicly reported and fueling an already active debate over AI safeguards within military and intelligence operations, further congressional or public scrutiny of how the Pentagon verifies AI-assisted intelligence reports before acting on them seems likely. For the full investigation, see the original CNN report.