Exploring the formation of learning burnout among college students in AI context: a serial mediation mechanism of AI dependence and addiction based on I-PACE model

Introduction The rapid evolution of artificial intelligence (AI) has transformed higher education by providing unprecedented support for AI-assisted learning. Despite these benefits, increasing concerns have emerged regarding students’ dependency on and addiction to AI technologies, which may contribute to learning burnout. Drawing on the I-PACE framework, this study proposes a serial mediation model to examine the psychological pathways linking technological affordances, AI dependence, AI addiction, and learning burnout among college students. Methods A cross-sectional questionnaire survey was conducted among 412 Chinese college students. The study measured perceived usefulness of AI, inert thinking, perceived enjoyment, AI dependence, AI addiction, and learning burnout. Structural equation modeling (SEM) and mediation analyses were performed using SmartPLS and SPSS to test direct, mediating, and serial mediation relationships among the variables. Results The findings indicated that perceived usefulness, inert thinking, and perceived enjoyment significantly and positively predicted both AI dependence and AI addiction. Furthermore, AI dependence and AI addiction were significant predictors of learning burnout. Serial mediation analysis revealed that AI dependence and AI addiction jointly mediated the relationships between perceived usefulness, perceived enjoyment, and learning burnout. In contrast, inert thinking influenced learning burnout indirectly through AI dependence alone rather than through the full serial pathway. Discussion This study elucidates a psychological mechanism through which technological affordances may evolve into dependency-related behaviors and subsequently contribute to learning fatigue in AI-assisted educational contexts. By extending the I-PACE model to higher education, the findings highlight the importance of fostering AI literacy and self-regulatory capacities to mitigate AI-related risks while preserving the benefits of intelligent learning environments. These insights provide theoretical and practical implications for promoting healthy AI use and sustainable learning experiences in the era of intelligent education.