Faculty are increasingly expected to integrate AI into teaching while preserving academic integrity and authentic student learning. This session offers a faculty-centered discussion of the lived classroom realities of student generative AI use—particularly the “AI confidence problem,” in which AI-generated writing can sound polished and authoritative while containing oversimplifications and missing nuance. Participants will reflect on common faculty challenges including uncertainty about student authorship and the emotional labor associated with trust and accountability in AI-enabled classrooms.Rather than framing AI as simply “allowed” or “banned,” this presentation proposes a values-driven approach to responsible AI use grounded in three core principles: transparency, verification, and student ownership of learning. The session shares a case study of redesigning an undergraduate writing assignment on culturally informed end-of-life communication. The redesign requires students to generate an AI response (Part 1) and then complete a structured credibility audit (Part 2) by identifying missed nuance or inaccuracies using five academic sources and a “claim → source → correction” method.
