Generative AI and student learning performance in medical higher education: a social cognitive theory perspective

Purpose: This study examines the impact of Generative Artificial Intelligence (GAI) on medical students' learning performance in higher education. Drawing on Social Cognitive Theory, it investigates the mediating roles of technology self-efficacy, perceived learning outcomes, and cognitive engagement, as well as the moderating role of fairness and ethical perceptions in AI-supported learning environments. Methods: A cross-sectional quantitative design was employed, with data collected from 326 medical higher education students in Pakistan using validated survey instruments. Structural equation modeling (SEM) with SmartPLS 4.0 was applied to test direct, mediating, and moderating relationships among the study variables. Results: The findings reveal that GAI-based technologies (GAT) exert a significant positive direct effect on students' learning performance and indirect effects through technology self-efficacy, perceived learning outcomes, and cognitive engagement. Technology self-efficacy and cognitive engagement emerged as strong mediators, while perceived learning outcomes further supported academic motivation and achievement. Moderation analysis demonstrated that perceptions of fairness and ethics significantly strengthen the positive effects of these mediators on learning performance, underscoring the importance of ethically grounded AI implementation. Conclusion: The study provides empirical evidence that GAI can enhance learning performance in medical higher education when supported by strong self-efficacy, engagement, and positive learning perceptions. Importantly, ethical and fair AI practices amplify these benefits. The findings offer practical implications for educators and policymakers seeking to integrate generative AI responsibly and effectively into medical education.