Teaching with Chatbots: Korean Language Instructors’ Perceptions of AI Integration and Instructional Value

ABSTRACT

This study investigates Korean language educators’ perceptions of artificial intelligence (AI)-powered chatbots as instructional tools in second and foreign language education. While previous research has emphasized learner outcomes, this mixed-methods study centers on the perspectives of 30 Korean language instructors across North America who engaged with three chatbot platforms—Sejong AI Tutor, Langotalk, and Eggbun. Participants evaluated the tools using surveys and detailed reflection reports, guided by the Chatbot-Human Interaction Satisfaction Model (CHISM) and the extended Technology Acceptance Model (TAM2).

Quantitative analysis revealed Langotalk as the most favorably rated platform, particularly in areas such as conversational realism, voice interaction, and user engagement. Sejong AI Tutor was valued for structured, beginner-friendly instruction, while Eggbun served as a useful review tool but lacked interactivity. Educators appreciated the motivational potential and accessibility of chatbots but expressed concerns about linguistic accuracy, honorific usage, error correction, and curricular alignment.

Qualitative findings highlighted the importance of positioning chatbots as supplementary—not primary—instructional tools. Educators emphasized the need for improved cultural and linguistic sensitivity, better onboarding, and professional development to support effective classroom integration. This study offers practical recommendations to developers, institutions, and educators for refining chatbot use in Korean language instruction and promoting a balanced, pedagogically sound approach to AI integration.