Pentagon AI tool falsely flagged Chinese ship, nearly sparked conflict

An artificial intelligence tool used by U.S. military analysts generated a false intelligence report that nearly led to an armed confrontation with China earlier this year, according to sources familiar with the incident.

A U.S. Special Operations Command Pacific analyst asked an AI chatbot to analyze the manifest of a Chinese vessel operating in the Middle East during the Iran conflict. The system combined open-source intelligence with classified signals intelligence and incorrectly concluded the ship was transporting components for a nuclear weapons program. The analyst then used AI to format these findings into a standard intelligence report that was distributed through military command channels.

Military response and last-minute discovery

Based on the faulty report, the U.S. military began preparing to intercept the vessel. Armed personnel were readying to board the ship, and military aircraft were already airborne when officials conducted a deeper review of the intelligence. Only then did they discover the report had been generated with AI assistance and that its conclusions were entirely false.

One source characterized the incident to CNN as having "almost started a war," noting that any U.S. assault on a Chinese vessel could have triggered retaliation and broader conflict.

Why it matters

The incident exposes critical risks as the Pentagon rapidly deploys AI across military operations without adequate safeguards. Defense Secretary Pete Hegseth has pushed since January to "unleash experimentation" and distribute AI tools to three million military personnel. But this near-miss demonstrates that AI systems can generate convincing yet completely fabricated intelligence that humans may not question—especially when formatted as standard reports. The technology's ability to fuse classified and open-source data makes its errors particularly dangerous, as analysts may assume the AI has access to information they cannot independently verify. A top Army cyber general recently warned that AI has caused a tenfold increase in vulnerabilities across Pentagon networks.

Both U.S. Special Operations Command Pacific and the Pentagon declined to comment on the incident when contacted.

The details were first reported by CNN's Katie Bo Lillis and Zachary Cohen.