Self-regulation and overreliance on artificial intelligence: Unpacking a paradox through a mixed-methods study in higher education

The rapid emergence of generative artificial intelligence (AI) in higher education has raised concerns about students’ overreliance on these tools for academic and personal tasks. Although generative AI can enhance productivity and creativity, excessive use may undermine key learning competences. In this context, self-regulation may function as a crucial protective factor. This study examined the relationship between self-regulation and overreliance on generative AI among university students. The sample consisted of 404 undergraduates (M Age = 20.78, SD = 4.12) enrolled in education-related degrees at the University of the Basque Country. Validated self-report scales measured dimensions of self-regulation and AI overreliance, and mediation analyses were combined with qualitative content analysis of open-ended responses. Results showed that goal setting significantly predicted perseverance, decision-making, and learning from mistakes. A paradoxical pattern emerged: goal setting was positively associated with greater AI overreliance, yet its indirect associations through perseverance and learning from mistakes were negative, suggesting an inverse relationship, indicating that stronger regulatory strategies are linked to lower levels of overdependence. These findings suggest that students who set clear goals tend to report higher initial use of AI tools, while perseverance and reflective learning are associated with patterns that counterbalance excessive reliance. Qualitative analyses supported this gradient, revealing that most students used AI minimally or for basic information searches, whereas a smaller subgroup relied on it for academic tasks or decision-making. This distribution indicates that overreliance is concentrated among a limited portion of students, consistent with the protective role played by strong self-regulatory competencies.