Antecedents and Consequences of Generative Artificial Intelligence Addiction Among Chinese College Students: A Self-Regulated Learning Perspective

The rapid diffusion of generative artificial intelligence (GAI) in higher education has generated both substantial learning benefits and growing concerns regarding students’ excessive and habitual use. Building on self-regulated learning theory, this study develops and tests an integrative research model comprising 10 hypotheses to examine the antecedents and consequences of GAI addiction among 1,022 Chinese college students. A self-administered online survey was conducted in China, and the data were analyzed using structural equation modeling (SEM). Results supported seven of the 10 hypotheses, revealing that GAI self-efficacy, GAI skill anxiety, peer pressure, and social identity were all significantly and positively associated with students’ GAI addiction. Unexpectedly, GAI replacement anxiety and need for cognition also positively related to GAI addiction, while GAI literacy showed no significant association. GAI addiction was further positively associated with creativity impact, teamwork performance impact, and decision reliance. Multi-group analysis further revealed that goal orientation significantly moderated several of these relationships, with cognitively driven pathways being stronger among low goal-oriented students, whereas identity-driven mechanisms were more salient among high goal-oriented students. This research contributes to the emerging literature on GAI-related behaviors by advancing a comprehensive, multi-factor perspective on GAI addiction and offering nuanced insights into its implications for higher-order learning outcomes in higher education.