國戰會論壇》滑鼠點三下 戰爭就開始了？（蔡裕明）
2026-04-17
中時新聞網
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
During the first day of US airstrikes on Iran, the Palantir-developed AI system Maven rapidly generated over 1,000 strike options by analyzing vast battlefield data. The AI's recommendations were used in real attacks, resulting in significant civilian casualties, including a school bombing, highlighting the risks of AI-driven military decision-making.[AI generated]
Why's our monitor labelling this an incident or hazard?
The article explicitly mentions an AI system (Maven Smart) used in military targeting that directly contributed to a lethal strike on a civilian target, causing significant loss of life. This is a clear case of harm to people (a) caused by the use of an AI system in a real-world context. The AI system's role in accelerating the kill chain and generating attack recommendations is central to the incident. The discussion of autonomous drone swarms and AI-driven information warfare further supports the presence of AI systems with potential for harm, but the realized lethal strike confirms this as an AI Incident rather than a hazard or complementary information. The article's detailed description of the event and its consequences meets the criteria for an AI Incident under the OECD framework.[AI generated]