Creative personal identity in the age of generative AI: A social-cognitive pathway of AI literacy, self-efficacy, and mindset

With the rapid integration of generative artificial intelligence (AI) into higher education, its influence on students' creative development has drawn growing attention. However, little research has systematically examined how AI literacy (AIL) shapes university students’ creative personal identity (CPI). To address this gap, this study developed a moderated mediation model grounded in identity theory and social cognitive theory, positioning AIL as the predictor, creative self-efficacy (CSE) as the mediator, and creative mindset as the moderator. Using a cross-sectional design, the study employed structural equation modeling and multi-group analysis. The results show that AIL significantly and positively predicts CPI, with CSE partially mediating this relationship. Moreover, fixed and growth mindsets exerted distinct moderating effects across the model's pathways. Multi-group analysis further revealed that among students with more than two years of exposure to generative AI, the impact of AIL on CPI was stronger, whereas the influence of CSE on CPI was weaker. From both cognitive and psychological perspectives, this study clarifies the mechanisms through which AIL shapes students’ creative identity and highlights how differences in AI usage experience alter these pathways. The findings provide a theoretical foundation for understanding creative identity construction in the digital era and offer actionable insights for designing curricula and teaching practices that foster creativity in university education.