The Impact of AI Algorithm Monitoring on Employee Job Satisfaction

Abstract

Purpose-Against the backdrop of artificial intelligence (AI) technology deeply permeating workplace management, this study focuses on the differential impact mechanism of AI algorithmic monitoring on employees' job satisfaction. Based on social cognitive theory and organizational identification theory, this study constructs and verifies a dual-driven path of developmental feedback and controlling feedback, exploring the mediating effects of self-efficacy and work identification. The aim is to provide a theoretical basis for achieving a balance between efficiency and humanistic care in digital management. Design/Methodology/Approach-A questionnaire survey was adopted to collect data from 343 employees in AI technology-intensive industries in eastern China. The partial least squares structural equation modeling (PLS-SEM) was employed to construct a "dual feedback-dual path mediation" model, thereby systematically testing the asymmetric impacts of developmental feedback and controlling feedback on job satisfaction, as well as their internal psychological mediating mechanisms. In the study, Harman's single-factor test was conducted to control common method bias, and the Bootstrap method was employed to verify the significance of mediating effects. Findings-Developmental feedback significantly enhances job satisfaction through its mediating effect on self-efficacy and job identity, following a positive feedback loop of “skill empowerment → efficacy construction → identity reinforcement.” In contrast, controlling feedback reduces job satisfaction by weakening self-efficacy and job identity, forming a negative feedback loop of “autonomy deprivation → identity dissolution.” The mediating effects of the two types of feedback exhibit significant differentiation in weight: in the developmental feedback pathway, self-efficacy plays a dominant role; in the controlling feedback pathway, job identity demonstrates stronger cultural sensitivity. Originality/Value-This study pioneers a dual-feedback classification system within algorithmic monitoring contexts, transcending the limitations of traditional one-dimensional monitoring research. It unveils the dual mediating mechanisms of competency development and organizational identity, clarifying a previously obscure “black box.” Findings confirm that the positive effects of developmental feedback significantly outweigh the negative impacts of controlling feedback. This provides critical theoretical support for designing AI management paradigms centered on developmental feedback with controlling feedback as a supplement, offering substantial practical guidance for cultivating employees' psychological resources during digital transformation.