Research centers on making AI understandable, trustworthy, and contestable by people. Combines social and behavioral science with machine learning to design, evaluate, and dialog about explanations; and advances AI planning and multi-agent reasoning for human-centered decision support.