AI-Assisted Decision Making

The field studies how people and AI systems can combine information and judgment to make better decisions in settings such as law, medicine, business, hiring, and public services. Approaches include decision-theoretic models of when to use AI advice, systems that retrieve evidence or coordinate specialized agents, and methods for explaining predictions, estimating uncertainty, or deferring to human reviewers. Researchers evaluate these systems with task benchmarks, simulations, field pilots, and experiments measuring both decision quality and how people use advice. A central tension is that accurate predictions do not guarantee better outcomes: advice can produce overreliance, bias, or misplaced confidence, so effective systems must also support human agency, accountability, and appropriate escalation.