The Good, the Bad, and the Ugly

Some predictive systems have helped governments improve inspections, anticipate service needs, target interventions, and allocate resources more effectively. Others have failed because of poor data, flawed assumptions, feedback loops, bias, or misplaced confidence in uncertain forecasts. Through real-world examples, participants will examine both successful and unsuccessful deployments of predictive AI and learn how to identify common warning signs, limitations, and unintended consequences before they become costly mistakes.