Xiaoyan Chen, Limin Xiao
As university students increasingly interact with AI, understanding how student-AI interaction behaviors are associated with creativity has gained increasing scholarly attention in recent years. However, previous research has yet to examine the correlation between human information behavior and creativity in AI usage, particularly in relation to information encountered with greater cognitive transformation potential. This study introduces the concept of AI information encounter in the context of student-AI interactions and explores its association with university students' creativity, including its mechanisms and boundary conditions, through the lens of Cognitive Flexibility Theory. Survey data were collected from 645 university students across different grades, regions, and majors. We complemented PROCESS with CB-SEM and PLS-SEM to triangulate the model, and the convergence across methods supports the model's stability. The results showed that AIIE positively predicted students' creativity, with cognitive flexibility serving as a positive mediator. Notably, the mediation of cognitive flexibility was only significant among students with medium to high levels of AI literacy, demonstrating a moderated mediation effect. The findings highlight the relevance of AI information encounters among university students and identify a mediating role linking AIIE to individual creativity, and shed light on practical implications for higher education institutions and teachers to cultivate university student creativity effectively.