Drivers and pathways of <scp>AI</scp> academic mentor acceptance: An <scp>SEM</scp> ‐ <scp>fsQCA</scp> study integrating cognitive appraisal theory and the <scp>AIDUA</scp> model

Abstract Artificial intelligence (AI) is reshaping learning in higher education, particularly within the global shift towards sustainable education and human‐centric visions. However, as traditional human mentoring faces challenges such as limited availability and inconsistent support, the potential of AI to function as an academic mentor remains underexplored. This study aims to investigate the alternative motivations for university students to accept AI as an academic mentor. Based on cognitive appraisal theory (CAT) and the artificially intelligent device use acceptance (AIDUA) model, we employed a mixed‐method approach combining structural equation modelling (SEM) and fuzzy‐set qualitative comparative analysis (fsQCA) to analyse the influencing paths and antecedent configurations. SEM results reveal that students' AI acceptance decisions exhibit a distinct three‐stage process: in the cognitive appraisal stage, perceived humanness and novelty value are the primary drivers; in the affective appraisal stage, performance expectancy is a stronger trigger for emotion than effort expectancy; and in the decision‐making stage, AI literacy emerges as the key determinant of final acceptance. fsQCA further identifies three typical configurations: an efficiency‐prioritized type driven by instrumental rationality, a social interaction type centred on emotional experience, and an exploration‐driven type characterized by the pursuit of innovation. These findings confirm that students' acceptance of AI academic mentors is not solely dependent on technical performance but is shaped by the complex interplay of cognitive, affective, and competency factors. The study provides important implications for higher education institutions seeking to integrate AI tools effectively and ethically.