When ease of use is not enough: self-efficacy and continued use of AI educational tools in resource-constrained universities

Abstract

Purpose Generative artificial intelligence (AI) is increasingly used in higher education, yet students’ continued use of AI educational tools may operate differently in resource-constrained university contexts. This study examines college students’ continuance intention to use AI educational tools in local undergraduate universities across five northwestern Chinese provinces and develops a capability-centric extension of the unified theory of acceptance and use of technology (UTAUT). Methods A sequential explanatory mixed-methods design was adopted. After pilot-based construct purification, 703 valid survey responses from students with prior AI tool experience were analyzed using partial least squares structural equation modeling (PLS-SEM). Additional analyses included common method bias diagnostics, PLSpredict, cross-validated predictive ability testing, multi-group analysis, and post-hoc power analysis. Semi-structured interviews with 25 students were used to contextualize the quantitative findings. Results Self-efficacy was the strongest predictor of continuance intention, followed by performance expectancy, perceived policy signals, and social influence. Algorithmic trust showed only a weak positive association, whereas effort expectancy showed a weak but significant negative association, contrary to the original UTAUT prediction. Qualitative findings indicated that students perceived AI tools as easy to access but cognitively demanding to use effectively, due to prompt refinement, factual verification, error correction, and task adaptation. Conclusion Continued AI educational tool use in resource-constrained universities depends less on surface-level ease of use and more on students’ perceived capability to manage the full AI-assisted learning workflow. This study extends UTAUT-based AI adoption research by foregrounding self-efficacy and proposing a capability-centric perspective on generative AI continuance in higher education.