From appropriate use to dependency: How AI attitude and use motivations shape university students’ AI dependency

Whether students develop dependency on generative artificial intelligence (AI) technologies largely depends on their attitudes toward using such tools. Existing studies have primarily explained the formation of AI dependency from the perspectives of usage attitude and technological characteristics, while paying limited attention to the psychological mechanisms through which attitudes exert their influence. Grounded in Self-Determination Theory, the present study examines how different types of use motivation mediate the relationship between students’ AI attitude and AI dependency. A total of 405 university students completed a comprehensive questionnaire measuring AI attitude, use motivations, and AI dependency. The results indicate that: (i) AI attitude does not directly predict AI dependency; instead, its effects are transmitted through multiple motivational pathways, with intrinsic motivation inhibiting AI dependency, whereas identified regulation facilitates its development; (ii) three distinct AI dependency profiles were identified: dependent, moderate, and conservative users. These profiles exhibited clear stratification, with dependent users showing the highest scores across all dimensions and conservative users the lowest; and (iii) individuals with stronger motivations are more likely to be classified into the dependent user profile, with external motivation playing a particularly salient role. These findings enrich the literature on AI dependency and offer practical implications for higher education administrators, tool developers, and educational researchers.