Pathways to Teacher Wellbeing: AI Pedagogy Self-Efficacy, Workload, and Anxiety in a Structural Model

As artificial intelligence (AI) tools become more common in K–12 settings, questions remain about whether AI competence can meaningfully support teachers’ psychological health. Grounded in Social Cognitive Theory, this study examined how teachers’ AI pedagogy efficacy relates to their instructional self-efficacy, engagement self-efficacy, workload, anxiety, and overall mental wellbeing. Using survey data from a nationally representative sample of 402 U.S. K-12 teachers, we employed hierarchical regression and structural equation modeling (SEM) to test direct and indirect pathways. Regression analyses indicated that AI self-efficacy, instructional self-efficacy, engagement self-efficacy, workload, and anxiety all significantly predicted teacher wellbeing. The SEM revealed an indirect pathway in which AI pedagogy efficacy strengthened teachers’ self-efficacy, which in turn reduced perceived workload, lowered anxiety, and ultimately improved teacher mental wellbeing. Findings highlight the importance of building teachers’ AI competence. As AI becomes embedded in K-12 instructional practice, supporting teachers’ confidence in using these tools may serve as a valuable lever for improving teacher wellbeing.