AI-Generated Intelligence Report Nearly Triggered US Strike on Chinese Vessel

A flawed intelligence assessment that included material generated by an artificial intelligence tool nearly prompted U.S. forces to strike a Chinese vessel in the Middle East, CNN reported, citing its own sources. A person who spoke to the network said the episode "almost started a war."

According to CNN, the report was assembled by an analyst at a special operations command who had queried an AI chatbot about the ship’s cargo manifest. The algorithm reportedly fused open-source material with classified electronic surveillance and concluded the vessel was carrying components related to a nuclear weapons program. Military planners were prepared to intercept the ship, but the operation was halted when officials determined a key section of the assessment had been produced by AI and had not been corroborated by human analysts.

Journalists have not been able to independently confirm what the ship was carrying, and the specific AI tool the analyst used—whether a commercial chatbot or a model adapted for government use—remains unknown. That ambiguity is partly a reflection of the Pentagon’s rapid embrace of AI across many areas. In January, Defense Secretary Pete Hegseth issued a memorandum instructing the department to adopt an "AI-first" approach and to test models from leading American companies.

Not all vendor relationships have proceeded smoothly. A planned arrangement with Anthropic collapsed after the company refused to provide its models for use in creating autonomous weapons; the government briefly blacklisted Anthropic and a court later ruled that ban illegal and unfounded.

Other partnerships have moved forward: in February SpaceXAI reportedly agreed to deploy Grok in classified military systems, and in May NVIDIA, Microsoft, and Amazon signed deals to supply AI technologies to the Department of Defense.

Those contracts can be highly lucrative and help explain the competition among suppliers. At the same time, the near-miss with the Chinese ship illustrates the particular dangers of relying on AI in military decision-making, where algorithmic errors can have consequences far beyond those typical for corporate software.