An investigation of the impact of generative AI's characteristics on construction project professionals' innovative work behavior

Purpose This study attempts to investigate the influence of generative artificial intelligence (GAI) on construction project professionals, specifically by exploring how GAI characteristics (intelligence, humanness, and unpredictability) impact innovative work behavior. Design/methodology/approach Grounded in the adaptive structuration theory for individuals, a theoretical model, with human–GAI collaboration as a mediator, was developed and empirically tested using the structural equation modeling approach. 226 valid responses from the Chinese construction industry, collected through a questionnaire survey, were used for the empirical analysis. Findings The findings reveal that GAI intelligence and humanness significantly promote human–GAI collaboration, which subsequently fosters innovative work behavior. Conversely, GAI unpredictability exerts a significant negative impact on collaboration. Furthermore, human–GAI collaboration effectively mediates the positive relationships between GAI's intelligence/humanness and innovative work behavior. Originality/value This study contributes to construction project management literature by shifting the analytical focus to micro-level individual behavioral pathways and distinguishing the differentiated impacts of GAI's relational characteristics with a robust empirical study.