How AI anxiety and confidence shape student engagement in higher education

The rapid integration of artificial intelligence (AI) into higher education offers opportunities for personalized learning and academic support but may also evoke anxiety, technostress, uncertainty, and concerns about competence, integrity, and future professional roles. This narrative review examines the relationships among AI anxiety, confidence, self-efficacy, and student engagement. A structured search of PubMed, Scopus, Web of Science, and Google Scholar identified relevant literature published between 2015 and 2026. Empirical studies, theoretical papers, and reviews addressing emotional responses to AI, student engagement, motivation, self-efficacy, and learning behavior were synthesized thematically using Self-Determination Theory, Self-Efficacy Theory, and emotion-regulation perspectives. The literature generally associates AI anxiety with avoidance, cognitive burden, reduced perceived control, and reluctance to experiment, whereas confidence and self-efficacy are linked to persistence, exploration, and active engagement. These relationships appear reciprocal and context dependent, shaped by prior technological experience, AI literacy, discipline, identity, cultural narratives, institutional policy, and the type of AI application. Excessive confidence may also encourage overreliance and uncritical acceptance of AI outputs. The review proposes a provisional five-stage framework comprising curiosity, anxiety, experimentation, confidence, and flourishing. This framework is intended as a heuristic rather than a universal or empirically validated sequence. Educational responses should combine scaffolded practice, transparent policies, critical AI literacy, mentorship, reflective dialogue, and human oversight. Longitudinal, experimental, qualitative, and mixed-method research is needed to test the proposed mechanisms and identify effective interventions.