Introduction The rapid adoption of artificial intelligence (AI) in higher education has increased college students’ reliance on AI tools. While AI enhances learning efficiency, it may also undermine key cognitive processes required for innovation. Methods Using survey data from 1,032 students, this study employed partial least squares structural equation modeling (PLS-SEM) and fuzzy-set qualitative comparative analysis (fsQCA) to examine how AI dependence, cognitive inertia, employment pressure, and academic utilitarian atmosphere shape students’ innovation capability. Results AI dependence significantly increases cognitive inertia, with cognitive dependence ( β = 0.570, p 0.001) exerting a stronger effect than tool dependence ( β = 0.161, p 0.001). Cognitive inertia reduces innovation capability ( β = −0.111, p 0.001) and serves as a key mediator linking AI dependence to innovation. Employment pressure strengthens the positive effect of AI dependence on cognitive inertia ( β = 0.045, p 0.05). A stronger academic utilitarian atmosphere further amplifies the negative impact of cognitive inertia on innovation capability ( β = 0.052, p 0.05). The fsQCA results reveal multiple pathways to high innovation capability, with low cognitive inertia emerging as a core condition across all effective configurations. Discussion This study clarifies the cognitive mechanisms and contextual conditions through which AI dependence affects innovation. The findings extend research in educational technology and innovation psychology and offer practical guidance for universities to optimize learning environments, promote rational AI use, ease employment pressure, and improve academic culture.
Fang et al. (2025) studied this question.