The role of AI adoption in the relationship between digital self-efficacy and sustainable entrepreneurship intentions among university students in Jordan

Purpose This study investigates associations among digital self-efficacy (DSE), artificial intelligence (AI) adoption intention and sustainable entrepreneurship intentions (SEIs) among university students at a private university in Amman, Jordan – a context with high unemployment and national digital agendas. Design/methodology/approach A sequential exploratory mixed-methods design was employed. Phase 1 involved semi-structured interviews with seven entrepreneurship faculty and an incubator director at a private university in Amman. Phase 2 comprised a survey of 150 students from the same institution. The technology acceptance model (TAM) and social cognitive theory (SCT) were integrated in a mediated model tested using partial least squares structural equation modelling. Findings DSE was positively associated with perceived usefulness (PU) (β = 0.570, p < 0.001) and perceived ease of use (PEOU) (β = 0.584, p < 0.001) of AI tools. AI adoption intention (AIAI) mediated the relationship between DSE and SEI. The total indirect effect of DSE on SEI through both serial mediation paths was 0.247 (p < 0.001). The specific serial indirect effects were 0.078 (DSE → PU → AIAI → SEI, p = 0.012) and 0.093 (DSE → PEOU → AIAI → SEI, p = 0.006). The model accounted for 42.8% and 25.3% of the variance in AIAI and SEI, respectively. These findings are limited to university students at one institution; replication with representative samples and longitudinal designs is needed before drawing broader conclusions. Practical implications Within university settings, educators might consider building DSE through experiential training that links AI to sustainability challenges. Universities could explore integrating AI literacy into entrepreneurship curricula. Originality/value This study positions AIAI as a potential mediating mechanism between DSE and SEIs, providing empirical evidence from a Middle Eastern university context. The contribution is incremental and context-specific, providing preliminary evidence consistent with the proposed theoretical mechanism.