Use of Predictive AI for Public Health Practice - Ian Painter, WA DOH - June 25, 2026

In his talk “You and What Data?” Differentiating Hype, Hope, and Reality in the Use of Predictive AI for Public Health Practice, Ian Painter, and infectious disease modeler with Washington Dept of Health, examines the type of AI commonly referred to as predictive AI, which largely includes methods previously described as machine learning. As the name suggests, the primary goal of predictive AI is to generate predictions. This distinguishes it from generative AI and agentic AI, though these approaches may share some underlying methods. Recent excitement, concern, and speculation surrounding large language models and agentic systems often spill over into discussions of predictive AI, sometimes creating unrealistic expectations about its capabilities. This presentation clarifies the similarities and differences between predictive AI and other forms of AI; reviews the current state of predictive AI; discusses practical applications in public health; and considers how to distinguish among what is currently feasible, what may become feasible in the near to medium term, and the barriers that may limit the utility of predictive AI in public health practice.