Comparing Performance Feedback from Human Managers to AI Managers: Effects on Employees‘ Perceptions, Basic Needs, and Work Engagement

Abstract

AI is increasingly implemented for managerial tasks, such as providing employee performance feedback. Yet, empirical research investigating employees’ perception and the impact of AI-assisted performance feedback on employees’ work engagement is scarce. We conducted two experimental studies and compared feedback from AI managers to human managers, drawing on self-determination theory. In both studies, we observed that feedback from AI managers is less accepted, perceived as lower in quality, and results in decreased relatedness satisfaction and work engagement compared to identical feedback provided by human managers. Moreover, we suggest three underlying mechanisms explaining how AI-assisted performance feedback diminishes employees’ work engagement. As our first study was conducted in 2022, shortly before the launch of ChatGPT, and our second, a direct replication, 1.5 years later in 2024, the results not only support the stability of these effects but also offer a unique perspective on temporal shifts in employees’ perceptions and the impact of AI-assisted performance feedback before and after the introduction of ChatGPT. Overall, our studies close current research gaps on the effects of delegating managerial tasks to AI on employees, focusing on performance feedback, and allow to derive concrete practical implications for the successful implementation of AI.

Ähnliche Arbeiten