Aircraft were in the air before anyone caught the error. A U.S. military operation against a Chinese vessel was aborted this spring after officials discovered the intelligence driving it had been hallucinated by an AI chatbot, CNN reported. The false report claimed the ship carried nuclear weapons components.

The failure had a specific shape, reported by TechCrunch. A Special Operations Command analyst queried a chatbot to synthesize open-source and classified signals intelligence. The tool misidentified the cargo manifest. The analyst then used the same chatbot to format the fabricated findings into an official-looking summary. That summary circulated across command channels unchallenged. A GovAI research scholar flagged the core problem: "It's important for service members to understand the uncertainty inherent to LLMs."

The gap to watch is not capability but verification. AI output moved up a command chain because it looked authoritative. No human reviewed the source before aircraft launched. Military buyers integrating AI for decision speed now face a harder question: what audit layer sits between model output and operational action.