Adoption of generative AI in academic writing among Chinese University students: an extended UTAUT model incorporating trust, moral obligation, and perceived risk

This study extends the Unified Theory of Acceptance and Use of Technology (UTAUT) by incorporating moral obligation, trust, and perceived risk to examine the adoption of generative artificial intelligence (GAI) for academic writing among Chinese university students. Data were collected through an online survey of 486 undergraduate, master’s, and doctoral students and analyzed using partial least squares structural equation modeling (PLS-SEM). The results show that trust was the strongest predictor of behavioral intention, whereas perceived risk had a direct negative effect on behavioral intention. However, the hypothesized link between trust and perceived risk, as well as the mediating role of perceived risk, was not supported. Performance expectancy and social influence positively affected behavioral intention, while moral obligation had a negative effect. Facilitating conditions positively predicted use behavior, whereas effort expectancy was not significant. Multigroup analysis showed no significant gender differences across comparable paths, but the negative effect of moral obligation on behavioral intention was stronger among doctoral students than among undergraduates. These findings advance current understandings of GAI adoption in academic writing, offering valuable implications for ethical governance and instructional support in higher education.