Purpose: This study aimed to identify and categorize factors influencing employee trust in AI-based decision-making within organizations and develop a conceptual framework that explains their relationships.
Research Methodology: A literature review combining bibliometric analysis with an in-depth examination of highly cited papers was conducted. A total of 136 articles from Scopus (2021–2025) were analyzed for this study. The ten most-cited papers were selected for thematic synthesis.
Results: Bibliometric mapping using VOSviewer identified publication growth, country distribution, subject areas, keyword co-occurrence, and author collaborations. The thematic synthesis of the ten most-cited papers grouped determinants of trust into AI system, individual, organizational, and task-context factors and emphasized calibrated trust as the core mechanism of trust-building.
Conclusions: Employee trust in AI is not static but a dynamic spectrum that requires calibrated trust. Self-confidence often outweighs trust in AI in determining the acceptance or rejection of AI recommendations.
Limitations: Reliance on Scopus-indexed articles (2021–2025) may have excluded historical perspectives. Qualitative synthesis has not been tested using meta-analysis. Sector and cultural heterogeneity limit the generalizability of the findings.
Contributions: Theoretically, this study integrates the Trust in Automation Model, the Integrative Model of Organizational Trust, and the SOR Model into a unified, dynamic, multilevel framework. Practically, it offers six recommendations for organizations to build sustainable employee trust in AI-assisted decision-making processes.