The adoption of AI in the workplace has sparked significant debates regarding its impact on employee outcomes. While some studies highlight AI’s potential to enhance employee capabilities and lead to positive outcomes, others caution that AI may displace human roles and contributions, resulting in negative consequences. Reconciling these conflicting perspectives is essential for harnessing AI’s benefits while mitigating its potential harms. This meta-analysis examines two distinct mechanisms—augmentation and automation—that explain how AI adoption influences employee outcomes. Our findings show that when both mechanisms are considered together, augmentation significantly boosts employee performance, well-being, and reduces deviant behaviors, while automation demonstrates less significant effects. However, moderator analyses reveal more pronounced automation effects among male employees, older workers, those interacting with virtual or advanced AI systems, and those in high-collectivism cultures. These results challenge the simplistic binary view that AI either augments or replaces human labor. Rather, we demonstrate that AI may primarily augment employee capacities, with its disruptive effects being less pervasive than often assumed but more pronounced in specific employee groups and contexts. Therefore, a more nuanced approach to the potential risks of AI integration is necessary, as exaggerated concerns may impede the full utilization of AI’s potential.
