Overview
- This spring a Special Operations analyst used a chatbot to examine a ship manifest and produced an AI-assisted intelligence report that falsely said the vessel carried material tied to a nuclear program.
- Military planners moved toward an interception with armed personnel preparing to board and aircraft airborne before officials traced the assessment back to the AI-generated analysis and halted the operation.
- Sources say the chatbot fused open-source information with classified signals intelligence and the analyst used AI again to format the findings into a standard intelligence report that circulated through military channels.
- It remains unclear which chatbot was used and what the ship actually carried, and neither the Pentagon nor US Special Operations Command Pacific has publicly clarified the episode.
- The incident highlights gaps in verification, tool provenance, human-in-the-loop rules and inconsistent standards across the Pentagon as it accelerates AI access, and it raises the risk of future miscalculations with China.