Advanced Legal Reasoning with Artificial Intelligence

The next generation of lawyers will practice co-intelligence—a dynamic partnership between human judgment and artificial intelligence in which their combined output exceeds what either could achieve alone. Taught by the Director of the BU Law AI Program, this seminar will teach you to work with generative AI as an active collaborator that can deepen your creativity, sharpen your reasoning, and expand your judgment on complex legal problems. Throughout the term, you will develop disciplined and reproducible methods for AI-enhanced legal work: decomposing novel problems into analyzable components, stress-testing arguments, running counterfactual and scenario analyses, and using AI to expose gaps or weaknesses that traditional workflows might overlook. You’ll gain a working understanding of large language models—their mechanisms, pattern-recognition strengths, and characteristic failure modes of hallucination, bias, and overconfidence—so you can exercise professional judgment about when AI output deserves acceptance, refinement, or rejection. Between class sessions, you’ll experiment with AI across legal use cases that demand creative and strategic thinking, then bring those explorations back to the seminar for collaborative critique. Together, students will compare approaches, analyze results, and co-build a shared playbook for high-level, AI-augmented lawyering. The course culminates in a “Before and After AI” capstone paper applying the cognitive partnership model to a hypothetical problem in litigation, regulatory advocacy, or transactional practice. UPPER-CLASS WRITING REQUIREMENT: Students may partially satisfy the requirement with a 3,000-word research paper. ** A student who fails to attend the initial meeting of a seminar or to obtain permission to be absent from either the instructor or the Registrar, may be administratively dropped from the seminar. Students who are on a wait list for a seminar are required to attend the first seminar meeting to be considered for enrollment.