With generative artificial intelligence (GenAI) increasingly embedded in hotel service processes, human–AI collaboration is shifting from simple automation to cognitive collaboration. Yet it remains unclear whether generative AI empowers employees or creates psychological challenges. Drawing on self-determination theory, social cognitive theory and regulatory focus theory, this study aims to develop a theoretical model examining the effects of GenAI–employee collaboration (employee-led vs GenAI-led) on hotel employees’ job satisfaction.
This study conducted two scenario-based experiments and one field experiment. Study 1 tested the main effect of GenAI-employee collaboration on job satisfaction and the mediating role of AI self-efficacy. Study 2 used a field experiment across three hotels with different star ratings to test the robustness and external validity of the findings. Study 3 adopted a 2 × 2 between-subjects design to examine the moderating role of work regulatory focus and the moderated mediation mechanism.
This study demonstrates that employee-led collaboration leads to higher job satisfaction than GenAI-led collaboration. AI self-efficacy mediates the relationship between collaboration type and job satisfaction. Work regulatory focus further moderates these effects. Promotion-focused employees report higher AI self-efficacy and job satisfaction under employee-led collaboration, whereas prevention-focused employees report higher AI self-efficacy and job satisfaction under GenAI-led collaboration.
Hotels should align generative AI workflows with task characteristics and employees’ motivational orientations. Employee-led collaboration is better suited to complex service tasks, while GenAI-led collaboration is better suited to standardized tasks. Managers should also enhance employees’ AI self-efficacy through scenario-based training and feedback.
This study shifts attention from whether generative AI is used to who leads GenAI-employee collaboration. It identifies AI self-efficacy as a key mechanism and work regulatory focus as a boundary condition. It also deepens understanding of human-technology fit in GenAI-employee collaboration.
