Using a Customized Generative AI Tool to Teach Feedback Skills for Navigating Difficult Leadership Conversations

Developing effective feedback skills is a persistent challenge in management education, as students often struggle with the truth, relational, and identity-based dynamics of feedback conversations. This article introduces an instructional innovation that integrates a customized generative AI tool, Feedback Coach, into an experiential learning exercise to support feedback skill development. Based on Kolb’s experimental learning framework, the exercise combines a case-based role play with Gen-AI supported practice that allows students to experiment with wording, tone, and goal alignment in a psychologically safe environment and revise their approach through structured reflection. Using The X-Caliber Project case, we describe the instructional design and classroom implementation and report evidence from pre-post qualitative reflections and survey data. Results indicate that students initially experienced misalignment, defensiveness, and uncertainty, while post-AI reflections and quantitative findings show increased confidence, clearer communication strategies, improved goal alignment, and higher perceived feedback effectiveness. We conclude with implications for instruction, boundary conditions, and guidance for adopting customized AI tools in experiential feedback training.