AI Information Encounters and University Students’ Digital Burnout: the Mediating Role of Cognitive Flexibility and the Moderating Effect of AI Learning Trust

Generative artificial intelligence (AI) has become an integral part of educational environments, where university students often encounter algorithm-recommended information unexpectedly, known as AI information encounters. Such interactions frequently lead to increased cognitive load, and it is unclear whether the ongoing depletion of these cognitive resources is statistically linked to digital burnout among university students. This study utilizes the Stimulus–Organism–Response (SOR) framework as its analytical foundation, employing cognitive load theory to elucidate the mediating role of cognitive flexibility, and social cognitive theory to explain the moderating role of AI learning trust, thereby constructing a moderated mediation model. A cross-sectional questionnaire survey, based on 566 valid responses, is conducted and analyzed using the PROCESS macro program (Model 4, 15) to examine mediation and moderated mediation effects. The results reveal that: (1) AI information encounters are significantly positively correlated with digital burnout; (2) cognitive flexibility partially mediates this relationship, as AI information encounters negatively correlate with cognitive flexibility, while cognitive flexibility negatively correlates with digital burnout; (3) AI learning trust negatively moderates the direct path between AI information encounters and digital burnout, meaning that higher levels of AI learning trust weaken their positive association; (4) AI learning trust also negatively moderates the link between cognitive flexibility and digital burnout, indicating that higher AI learning trust diminishes the negative association between cognitive flexibility and digital burnout. This study highlights the psychological costs associated with involuntary technology use and suggests that educators in higher education should be cognizant of the cognitive resource consumption resulting from AI information encounters, and the interactive effects of AI learning trust on this relationship. Additionally, educators should encourage students to develop reasonable AI learning trust and acknowledge the risk of digital burnout linked to AI information encounters.