From Confidence to Commitment: Self-Perceived Competence and the Motivational Pathways to Intellectual Engagement with AI Learning Tools

Artificial intelligence (AI) tools are becoming part of how students think, not only what they use. Whether a learner engages with such a tool or delegates to it is a decision about how their own cognitive resources are deployed, which places that decision within the realm of study of human intelligence. Investment accounts of intellectual development hold that ability is built through motivated engagement with cognitively demanding material, yet technology acceptance frameworks emphasize utility judgments and underrepresent the motivational processes that drive such engagement. This study tests an extended Technology Acceptance Model in which self-efficacy shapes perceived usefulness through motivation and enjoyment. Survey data from 185 undergraduate School of Business students at a Saudi university who used generative AI tools during coursework were analyzed using partial least squares structural equation modeling. All nine hypotheses were supported. Self-efficacy strongly predicted both motivation and enjoyment, but motivation was the dominant mediating pathway to perceived usefulness (β = 0.295, p < 0.001), with enjoyment a weaker affective route (β = 0.086, p < 0.05), alongside a significant serial pathway from self-efficacy through motivation to enjoyment (β = 0.113, p < 0.001). The model explained 54.4% of the variance in behavioral intention. The findings indicate that beliefs about one’s own technical competence, rather than appraisal of the tool alone, govern how students commit to intelligent systems, with implications for whether AI use develops or displaces learners’ own intellectual capacity.