Mirrors with Teeth: How AI Simulations Use Productive Failure and Personalized Feedback to Develop Adaptive Leadership

This mixed-methods study examines the pedagogical efficacy of AI-driven leadership simulations in undergraduate leadership education. The author served as course instructor. Twenty-nine students completed six sequential simulations paired with a reflective writing assignment over one six-week summer term. Qualitative thematic analysis reveals three primary outcomes: (1) simulations produced characteristic dip-then-climb learning trajectories consistent with productive failure frameworks, (2) personalized career profiles functioned as "mirrors with teeth" surfacing gaps between self-perception and demonstrated behavior, and (3) students underwent epistemological shifts from certainty-seeking to ambiguity tolerance. Quantitative self-report data indicates significant increases in adaptive leadership orientation (17% to 62%), listening-first approaches (24% to 76%), and comfort with uncertainty (21% to 66%). Findings suggest that consequence-driven simulation environments, when paired with individualized feedback and structured reflection, can accelerate the transition from declarative leadership knowledge to enacted judgment, particularly by surfacing relational skill gaps among analytically strong students.