What are the best practices for using generative AI in the legal process? And how may they change in the years ahead? This panel considers some answers in this important and fast-moving area, including: creative ways to use generative AI to improve the analysis of precedent and case records; the continuing challenge from hallucinated case and record citations; the reliability of AI-generated evidence; issues of privilege and confidentiality that arise from sharing information with AI models; and— most importantly—the role of human judgment in a world where AI serves as a pervasive assistant to clients, counsel, and the courts. Other questions include the disclosure of AI use and how AI tools can best be used without compromising fairness or public confidence.