Employee–AI Collaboration and Innovative Work Behavior in the Public Sector: A Mixed PLS-SEM and fsQCA Study

AI is rapidly reshaping public sector work and influencing employees’ innovative behavior. Drawing on the Job Demands–Resources model and the cognitive appraisal theory of stress, this study develops a moderated mediation model to examine the relationship between employee–AI collaboration and innovative work behavior in the public sector, focusing on the mediating role of creative self-efficacy and the moderating role of AI anxiety. Using survey data from 362 Chinese public-sector employees, this study combines PLS-SEM and fsQCA. The PLS-SEM results show that task support collaboration is significantly and positively associated with public-sector employees’ IWB, with CSE playing a significant mediating role in this relationship. In contrast, neither the direct association between social support collaboration and IWB nor its indirect association through CSE is statistically significant. The moderating role of AI anxiety varies across different collaboration pathways. The fsQCA results show that high IWB is not associated with any single factor alone but rather with different combinations of employee–AI collaboration forms, CSE, and AI anxiety, revealing multiple configurational pathways associated with high IWB. This paper advances research on employee–AI collaboration and IWB in the public sector, while offering practical guidance for organizations seeking to optimize collaboration models and unlock employees’ innovative potential.