Driverless metros are coming, but what about the drivers? A study on AI-related anxiety and safety performance

The application of automatic train operation (ATO) systems in rail transit gradually replaces manual labor, greatly impacting metro drivers' daily work. Building on the cognitive appraisal theory of stress and the socially embedded model of thriving at work, this study explored how drivers' perception of ATO systems would affect their safety behavior within the workplace, in other words, the psychological and behavioral costs of drivers' AI-related anxiety. Specifically, we analyzed whether job insecurity and thriving at work would mediate the association between AI-related anxiety and safety performance. We also examined the potential attenuating effects of Human Resource (HR) system strength on AI-related anxiety's impact on job insecurity, thriving at work, and safety performance. We gathered 364 sets of paired data from Chinese metro drivers and their direct supervisors based on a time-lagged design. Moderated serial mediation analyses revealed that AI-related anxiety exerted a significant serial indirect impact on safety performance via job insecurity and thriving at work. Additionally, we found conditional serial indirect effects of AI-related anxiety on safety performance as a function of drivers' perceptions of HR system strength. Theoretical and practical implications are discussed for a scholarly and practical understanding of dealing with AI-related anxiety. Metro organizations were recommended to place more emphasis on the HR process to mitigate the costs of drivers' AI-related anxiety.