[Survey and Practical Report] Using AI Prompts to Develop University Students' English Speaking Skills: Learners' Autonomous Learning Supported by AI and the Instructor

This practical report analyzes the outcomes of English speaking instruction using AI prompts implemented in the elective course Active Listening I during the first semester of 2025. Nineteen students engaged in English conversation practice over 13 weeks using nine types of AI prompts developed by the author. A five-point Likert scale questionnaire showed high evaluations, with an average score of 4.8 (out of 5) for the statement, “AI conversation was helpful for improving English proficiency.” All questions received high evaluations (average 4.4–5.0). Qualitative analysis of 226 AI conversation reports revealed that learners progressed through four stages: initial, improvement, plateau, and development. In addition, the learning processes of four learner types (eight students), classified according to AI conversation time, were analyzed. This practice demonstrated that when teachers respect learners' self-regulated learning while providing appropriate support, learners' au-tonomous and exploratory use of AI enhances their English speaking skills.