AI in Supply Chain Risk Assessment: A Systematic Literature Review and Bibliometric Analysis

AI significantly improves supply chain risk prediction: Machine learning models like Random Forest and XGBoost demonstrate substantially better accuracy in predicting supply chain risks compared to traditional methods, with hybrid approaches showing particular promise for adapting to post-pandemic disruptions.

Explainable AI is essential for responsible implementation: The research emphasizes that explainable AI (XAI) must be integrated into supply chain risk systems to ensure transparent, understandable decision-making—a critical requirement for organizations and stakeholders to trust and validate AI recommendations affecting business operations.

Data quality and interpretability remain major challenges: Despite AI's potential, the study identifies persistent obstacles including data quality issues and model interpretability problems that organizations must address to reliably deploy these systems in real-world supply chain environments.

Real-time data and blockchain enhance accountability: The research highlights that combining AI with real-time data utilization and blockchain technology for traceability creates more robust and auditable supply chain risk management systems, supporting both operational resilience and transparency.

Interdisciplinary collaboration is necessary for success: The findings stress that optimizing AI-driven supply chain risk assessment requires ongoing collaboration across disciplines and continuous model evaluation, rather than one-time implementations, to keep pace with evolving risk landscapes.