The impact of risk perception on perceived benefits of generative artificial intelligence users-an empirical study based on university students.

Amid the rapid integration of generative artificial intelligence (Gen AI) into higher education, it remains unclear how the risk perception ability of university students, as a core group of users, is associated with their perceived benefits of Gen AI. To address this gap, this study develops an integrated theoretical framework drawing upon Cognitive Appraisal Theory (CAT), Protection Motivation Theory (PMT), and the Technology Acceptance Model (TAM)/UTAUT2 to explain how different dimensions of risk perception differentially influence distinct categories of perceived benefits. Questionnaire surveys were conducted to collect valid data from 814 university students. Hierarchical regression analysis using SPSS was employed to examine the effects of demographic variables (gender, age, grade), Gen AI usage duration, and risk perception on perceived benefits. The results indicate that university students' risk perception of Gen AI is multidimensional, and different dimensions show significantly differential associations with perceived benefits. Specifically, security risk perception has a significant positive effect on academic assistance and skill development. In contrast, information risk perception exerts significant negative effects on psychological and emotional support, daily life, and leisure and entertainment benefits. Results of the threshold effect show that when college students' risk perception exceeds a certain threshold, the effect turns positive and significant, enhancing their perceived benefit. Based on these findings, it is recommended that related groups take actions to promote the development and application of Gen AI.