The Double-Edged Effects of AI: A Dual-Path Systems Perspective Based on Fairness and Dehumanization

In the context of AI being deeply embedded in organizational operations, employee work engagement has become a critical mechanism for translating technological potential into realized value. However, existing research offers inconsistent findings on how enterprise AI adoption influences work engagement, highlighting the need for theoretical integration. Drawing on Conservation of Resources and Social Information Processing theories, this study conceptualizes AI adoption as a resource-restructuring mechanism that shapes employees’ resource environments. A dual-path model is proposed, in which AI adoption affects work engagement through a resource gain pathway (fairness perception) and a resource loss pathway (perceived organizational dehumanization). Using three-wave time-lagged data from employees in knowledge-intensive and highly digitalized enterprises in China, the results show that AI adoption simultaneously enhances work engagement via fairness perception and reduces it via perceived dehumanization. Furthermore, AI transparency serves as a key system-level moderator that strengthens the positive pathway while weakening the negative pathway. By integrating resource gain and loss mechanisms, this study provides a system-level explanation of AI’s dual effects and offers insights for balancing technological efficiency with human-centered values in digital transformation.