Student technology self-efficacy statistically mediates the associations of institutional and online instructor support with AI literacy in AI-mediated online learning

The rapid integration of artificial intelligence (AI) into education has transformed how students learn and interact with knowledge, making AI literacy an essential competency in the digital age. Grounded in Social Cognitive Theory, this study examined the associations of institutional support and online instructors’ support with AI literacy and the mediating role of student technology self-efficacy among Chinese university students in AI-mediated online learning environments. AI literacy was conceptualised as a multidimensional construct comprising Technical Understanding, Critical Appraisal, and Practical Application. A cross-sectional survey was conducted among 603 Chinese university students aged 18–26 years in Shanxi Province, China. Data were analyzed using structural equation modeling (SEM). Institutional support and online instructors’ support significantly predicted both student technology self-efficacy and AI literacy. The standardised association between online instructors’ support and AI literacy was numerically larger than that between institutional support and AI literacy. Student technology self-efficacy partially mediated the relationships between institutional support, online instructors’ support, and AI literacy. These findings support the application of an SCT-informed framework to AI-mediated learning environments, showing that environmental support is statistically associated with AI literacy, both directly and indirectly through student technology self-efficacy. The study contributes to AI education research by clarifying the statistical relationships through which institutional support and online instructors’ support are associated with AI literacy.