
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
The U.S. National Weather Service used generative AI to create weather maps that included fictional town names in Idaho, such as "Whata Bod" and "Orangeotild." The incident, which led to public misinformation and eroded trust, highlights the risks of unsupervised AI in critical government communications. The errors were later corrected.[AI generated]
Why's our monitor labelling this an incident or hazard?
The AI system (a generative AI tool used to create weather maps) malfunctioned by hallucinating fictional town names, leading to the dissemination of false information by a critical government agency. This misinformation can erode public trust and potentially endanger lives if people rely on inaccurate forecasts during emergencies. The harm is indirect but significant, affecting the community's right to accurate information and safety. The event involves the use and malfunction of an AI system causing realized harm, fitting the definition of an AI Incident rather than a hazard or complementary information.[AI generated]