Employee acceptance of AI-driven performance management systems in hospitality

Artificial intelligence (AI) is increasingly used to evaluate employee performance in the hospitality industry, yet its relationship with employees’ fairness perceptions remains poorly understood, particularly when algorithmic decision processes are only partially observable. Challenging the direct applicability of organisational justice theory, this study proposes and tests a trust-mediated model in which algorithmic opacity shifts fairness judgements from procedural evaluation towards trust-based inference. Data were collected from 429 employees working in Chinese four- and five-star hotels with direct experience of AI-based performance management systems. A hybrid analytical approach combined partial least squares structural equation modelling (PLS-SEM), to test the proposed linear relationships, with an artificial neural network (ANN) analysis, to assess the relative predictive importance of the model’s predictors. The results show that AI applications are positively associated with employees’ perceived fairness of performance assessments, and that trust plays a central mediating role in this relationship, indicating that fairness judgements are largely inferred through trust rather than assessed directly through observable procedures. Algorithmic transparency strengthens the association between AI applications and both trust and perceived fairness, although its influence diminishes once trust is established. The ANN results complement the PLS-SEM findings by confirming the relative predictive importance of trust, AI applications, and algorithmic transparency, without indicating specific non-linear functional forms. By integrating trust, transparency, and predictive analysis, this study extends organisational justice theory to AI-mediated performance management and offers theoretical and practical insight for managing AI adoption in hospitality and service organisations.