Exploring the Relationship Between Generative AI and Employee Creativity in Construction Firms: A Hybrid PLS-SEM, IPMA, and fsQCA Approach

Background: This study focuses on construction firms undergoing digital transformation, exploring the mechanisms through which Generative AI (GenAI) is associated with employee creativity in a context where knowledge is highly context-dependent, project teams are temporary, and unique safety and schedule pressures prevail. Methods: A mixed-methods approach integrating Partial Least Squares Structural Equation Modeling (PLS-SEM), Importance-Performance Mapping Analysis (IPMA), and Fuzzy Set Qualitative Comparative Analysis (fsQCA) is employed. The proposed model is tested using primary survey data from 268 employees of Chinese construction firms. Results: Generative AI has no significant direct association with construction firm employee creativity (CFEC). Instead, it shows an indirect association through the full mediation of explicit knowledge sharing (EKS) and tacit knowledge sharing (TKS), with TKS having a stronger association with employee creativity. The relationship between GenAI and knowledge sharing is positively moderated by digital self-efficacy. The fsQCA identifies seven equifinal configurations leading to high employee creativity, with ‘explicit knowledge sharing and digital self-efficacy’ constituting the optimal configuration. Conclusions: Construction firms should actively promote knowledge sharing among their staff and provide regular training on GenAI tools, thereby fully harnessing employee creativity. Managerial Implication: Construction CEOs should prioritize building GenAI-supported knowledge sharing systems and improving employees’ digital self-efficacy, rather than expecting direct creativity improvement from GenAI deployment.