Discovering the Dark Side of AI and Algorithmic HRM: Evidence from Negative Experiences of Employees on Online Delivery Platforms

As employees are important stakeholders within the organization, it is crucial for enterprises to value employee voice behavior (VB). However, with the rise of algorithm management, whether and how this AI-driven paradigm impacts this important VB remains poorly understood. This study aims to uncover the dark side of algorithmic management by investigating the inhibiting effect of employees’ negative algorithmic experiences on their VB and its underlying mechanism. Applying Conservation of Resources (COR) theory as our framework, we leverage a big data text analysis approach, conducting text mining and Structural Topic Modeling (STM) on a large corpus of employee reviews for online delivery platforms sourced from Glassdoor. The findings show that employees’ negative algorithm experiences are mainly derived from AI algorithm matching and AI algorithm control. Subsequently, the negative algorithm experiences diminish employees’ recognition of organizational culture, which, in turn, suppresses their VB. Furthermore, the presence of work-life balance does not alleviate the inhibitory effect that negative algorithmic experiences have on employee VB. Through empirical analysis, this study reveals a negative relationship between algorithmic management and VB. These findings offer important theoretical and practical implications for gig platforms enterprises to optimize their algorithmic design and strengthen cultural identification, thereby encouraging employee voice.