TL;DR
A US military analyst used an AI chatbot to produce an intelligence report falsely claiming a Chinese ship carried nuclear weapons components, nearly triggering an armed boarding operation that could have sparked war with China.
What happened
- This spring, during active US military operations against Iran, a Special Operations Command Pacific analyst queried a chatbot about a ship's manifest and received a fabricated nuclear-weapons finding.
- The AI fused open-source intelligence with classified signals intelligence and produced a conclusion that was, per one source, "entirely false."
- The analyst then used AI a second time to format the findings into a standard intelligence report, lending it institutional credibility before wide circulation.
- Armed US personnel were preparing to board the Chinese vessel and military aircraft were already airborne before officials caught the error.
- One source said the report "almost started a war" between the United States and China.
Why it matters
- No unified verification standard exists across the military and intelligence community for AI-generated outputs, with different units using different tools under different safety rules.
- The chatbot involved was likely a near-copy of a commercial product: a former senior official said internal tools are "mostly just copies of the commercial stuff wearing lipstick."
- Defense Secretary Pete Hegseth's January 2026 AI Acceleration Strategy explicitly pushes AI into the hands of all three million military and civilian personnel at all classification levels, accelerating the risk surface.
- AI hallucinations in intelligence work are not isolated, according to sources: the proliferation of these tools has made such errors a recurring pattern, not a one-off.
- Young analysts in particular are flagged as trusting AI outputs uncritically, and institutional pressure to produce intelligence faster compounds the hazard.
- One source summarized the core danger: "AI allows you to get to a bad idea faster."
What to watch next
- Whether the Pentagon issues binding verification protocols for AI-generated intelligence reports, especially those feeding targeting decisions.
- Congressional or inspector-general scrutiny of US Special Operations Command Pacific's AI tool procurement and whether the chatbot was commercial or government-built.
- Any signal that China learned of the incident, which would carry its own escalation and diplomatic fallout independent of the near-boarding itself.
Originally published on Present of AI, a daily source-linked AI news timeline. Read the full timeline or browse the open dataset.