The behavioral determinants of generative AI use and self-directed learning outcomes in higher education

Generative artificial intelligence is rapidly reshaping how individuals learn and interact with knowledge. However, little is known about how learners perceive these tools and how such perceptions influence their ability to direct their own learning. Drawing on Garrison’s [17] framework of self-directed learning, the Technology Acceptance Model, and constructs of trust and risk, this study analyzes data from 228 business and technology students at a public university in the southwestern United States. Findings indicate that perceived usefulness, ease of use, and trust positively influence students’ adoption of generative artificial intelligence, while perceived risk exerts an indirect effect through its impact on trust. Intention to use artificial intelligence strongly predicts motivation and self-monitoring, and to a lesser extent, self-management. These results extend the Technology Acceptance Model by incorporating trust and perceived risk as socio-cognitive factors shaping generative AI adoption and by demonstrating how behavioral intention connects technology beliefs to self-directed learning outcomes. The findings contribute to research on educational technology adoption by clarifying how generative artificial intelligence can support learner autonomy, motivation, and self-regulation in higher-education learning environments.