The Impact of AI on Korean Language Learning Performance: An Empirical Study from the Perspective of Technological Mediation

AI technology is reshaping language learning by enhancing personalization and feedback, yet its impact on Korean language learning and underlying mechanisms remains insufficiently studied. This research adopts the perspective of technological mediation and constructs a model of AI application, intelligent recommendation system, and learning performance. Using questionnaire and academic data from Korean-major undergraduates, multiple regression and mediation analyses reveal that AI significantly improves performance (β=4.684, p<0.01). The intelligent recommendation system partially mediates this effect as AI usage increases recommendation system engagement (β=3.127, p<0.01), which further enhances learning outcomes (β=2.827, p<0.01). Marginal effect analysis shows optimal improvement at moderate AI use (β=5.127), with diminishing returns at higher levels. These results demonstrate that AI not only directly boosts academic performance but also indirectly enhances it by improving learning resource alignment. The study confirms the mediating role of intelligent systems and offers theoretical and empirical insights for optimizing AI-assisted language learning and curriculum design.