ChatGPT as a Study Partner : An Extended TAM Approach to Self-directed Learning Adoption

ChatGPT is increasingly integrated into daily and educational contexts, yet most uses remain short-term and task-focused. This study explores the factors that encourage learners to adopt ChatGPT for self-directed learning (SDL). Extending the Technology Acceptance Model (TAM), we include personal factors (self-efficacy, learning autonomy, learning motivation) and technological attributes (interactivity, compatibility, relative advantage, observability), examining their influence on satisfaction and intention to continue use. Data were collected from 467 users in Korea with prior ChatGPT experience, and analyzed through Covariance-Based Structural Equation Modeling (CB-SEM). Results show that self-efficacy, motivation, autonomy, interactivity, and relative advantage significantly affect perceived ease of use and usefulness. Perceived usefulness strongly predicts satisfaction and intention to use, with satisfaction mediating part of this relationship. The findings extend TAM by integrating personal and technological dimensions, offering insights for educators and developers on how AI tools like ChatGPT can effectively support sustained and autonomous learning.