대학 LMS의 AI 챗봇의 한계와 교육적 활용 가능성 = AI Chatbots in University LMS: Educational Potentials and Limitations

다국어 초록 (Multilingual Abstract)

This study examines the actual functions and educational implications of AI chatbots integrated into university Learning Management Systems (LMS), while exploring their potential and limitations from the user experience perspective. AI chatbots from seven Korean universities were selected, and their characteristics were compared across functional, technical, and educational dimensions based on official documents, user manuals, and demonstration videos. In addition, in-depth interviews were conducted with six undergraduate and graduate students and three faculty members from one selected university, and the data were analyzed thematically.
The findings reveal a clear discrepancy between the designation of “AI chatbots” and their actual functions within LMS. Although some universities demonstrated the potential for adopting advanced functions based on Retrieval-Augmented Generation (RAG) or Large Language Models (LLMs), the majority of chatbots primarily supported basic administrative tasks. As a result, advanced learning support features―such as contextual understanding and personalized feedback―were minimally implemented. Furthermore, users from the interviewed university highlighted common issues, including limited integration with course content, formulaic response patterns, and low awareness of the AI chatbot’s existence. Overall, the study provides empirical evidence of a substantial gap between the technical potential of LMS-based AI chatbots and users’ actual experiences. To move beyond simple functional adoption, the study suggests that development should emphasize functional sophistication informed by user experience, system stabilization, and effectiveness validation.