AI-generated false report nearly triggered US-China military clash, CNN reports

The US military came dangerously close to a confrontation with China this spring after an artificial intelligence chatbot produced a false intelligence report claiming a Chinese vessel in the Middle East was carrying nuclear weapons components, CNN reported Friday in an exclusive investigation.cnn

Armed military personnel were preparing to board the ship and military aircraft were already airborne before officials took a closer look at the report and discovered the AI-generated assessment was, in the words of one source, "completely erroneous" and had "almost initiated a war".mezha+1

How AI Produced a False Conclusion

The incident began when a special operations command analyst uploaded intelligence about the ship's manifest to an AI chatbot. The information had been obtained by US Special Operations Command Pacific, based in Hawaii. The chatbot combined open-source intelligence with classified signals intelligence from government databases and reached an incorrect conclusion about the vessel's cargo.mezha+1

The analyst then used AI again to format the findings into what appeared to be a standard intelligence report — the kind military officials typically trust without question — and distributed it through military channels. CNN could not determine what the ship was actually carrying.cnn

"A lot of the internal tools are merely adaptations of commercial software, dressed up for military use," a former senior official familiar with the AI systems used by the military told CNN.mezha+1

Growing Reliance, Growing Risk

The near-miss comes as the Pentagon accelerates its adoption of AI across defense operations. In January, Defense Secretary Pete Hegseth unveiled the department's "Artificial Intelligence Acceleration Strategy," aimed at making AI tools available to roughly three million civilian and military personnel across all classification levels.investing+1

Yet the effort remains fragmented, according to several US officials. Different sectors of the government employ various tools under distinct safety protocols, with no unified standard for verifying AI-generated information.cnn+1

Sources told CNN the military is increasingly using AI in targeting decisions, raising the risk of mistaken strikes, civilian casualties, and friendly fire incidents — even when a human formally retains final authority.mezha+1

A Warning From Within

The type of AI "hallucination" that nearly triggered the China incident has not been an isolated occurrence within the intelligence community since these tools began to proliferate, according to one source. Younger analysts, who are more comfortable with AI tools, may be especially inclined to accept their outputs uncritically, sources said.cnn

"AI enables you to arrive at a poor conclusion more quickly," one source remarked.mezha+1