Belief to behavior: how AI self-efficacy shapes AI acceptance through AI literacy

The increasing integration of artificial intelligence (AI) technologies into higher education has raised important questions regarding students’ acceptance of AI and the psychological and cognitive factors shaping their engagement with these systems. This study aimed to examine how university students’ AI self-efficacy influences their acceptance of AI, whether AI literacy helps explain this relationship, and whether greater AI acceptance is associated with lower AI anxiety. For the present study, data were obtained from 622 undergraduate students at a public university. Empirical analyses were based on validated scales measuring AI self-efficacy, AI literacy, AI acceptance, and AI anxiety, and conducted using correlation analyses and regression-based mediation models via the PROCESS Macro. The results showed that AI self-efficacy was positively associated with AI acceptance, both directly and indirectly through AI literacy. Higher AI acceptance was also associated with lower AI anxiety; however, this relationship should not be interpreted as causal. These findings support a sequential motivational, cognitive and affective model of AI engagement and highlight the importance of integrating confidence-building interventions with literacy-focused instructional strategies to promote psychologically adaptive and sustainable AI integration in higher education contexts.