Employee emotions and emotional labor after robot service failure: Anthropomorphism as moderator

The tourism and hospitality sectors are rapidly deploying AI and robotic technologies into service delivery, where employees work with robots to deliver operational benefits but still face robot inevitable errors (robot service failure). Correspondingly, research on robot service failure predominantly focuses on the cognitive processing to investigate its impacts on employees. Drawing on cognitive appraisal theory of emotion, this research extends the literature by investigating how employees respond to robot service failure through the affective processing. We propose that robot service failure can evoke three distinct employee emotions—anger, anxiety, and guilt—with countervailing effects on employees’ emotional labor (surface acting and deep acting). Specifically, robot service failure elicits anger and anxiety, which enhance surface acting and reduce deep acting; but also guilt, which enhances deep acting while not necessarily reducing surface acting. Furthermore, we extend the cognitive appraisal theory of emotion by incorporating robot anthropomorphism as a moderator shaping these effects. We test our model through two studies: Study 1, an experiment with tourism and hospitality employees in US, and Study 2, a multilevel experience-sampling field survey with Chinese hotel employees. This research offers theoretical and practical insights into robot service failure management.