Quiet Prompting under Algorithmic Governance: How Generative AI-Enabled HRM and Economic Bargaining Power Shape Discretionary Effort

Abstract

We examine how generative AI–enabled human resource management (HRM) systems shape a subtle form of withdrawal that we call quiet prompting. Quiet prompting refers to employees’ selective reduction of discretionary effort in response to AI‐driven HR decisions. Integrating organizational justice, social exchange, and algorithmic governance perspectives, we argue that generative AI–enabled HRM (GenAI-HRM) influences quiet prompting through employees’ perceptions of algorithmic fairness, algorithmic transparency, and trust in AI-driven HRM. We further theorize that perceived economic bargaining power conditions the effect of trust on quiet prompting, such that employees who feel less able to exit their organization are especially sensitive to how they are treated by AI systems. We test our model using a three-wave time-lagged survey of 634 employees in 11 countries whose organizations had implemented at least one generative AI–based HR module. Structural equation modeling shows that GenAI-HRM is positively associated with perceived algorithmic fairness and transparency, which in turn build trust in AI-driven HRM. Higher trust is linked to lower quiet prompting, and this negative relationship is significantly stronger for employees who report low economic bargaining power. Our findings introduce quiet prompting as an AI-specific form of selective withdrawal, extend justice and trust theories to generative AI–enabled HRM, and highlight how perceived labor-market power shapes employees’ behavioural responses to algorithmic governance.