Hybrid AI and Multi-Criteria Decision-Making for Predicting and Managing Supply Chain Disruptions: A Systematic Literature Review

Supply chains are increasingly exposed to disruptions caused by pandemics, geopolitical shocks, natural hazards, labour shortages, and infrastructure failures. This paper
investigates how artificial intelligence (AI), and multi-criteria decision-making (MCDM) are being combined to predict, assess, and manage such disruptions. The study adopts a systematic literature review approach, guided by PRISMA principles, to examine research published between 2020 and 2025 on hybrid AI-MCDM approaches for supply chain disruption management. The review analyses the AI techniques used, the MCDM methods applied, their main application areas, the types of data employed, and the evaluation measures reported. The findings indicate a growing shift towards hybrid decision-support frameworks in which machine learning and deep learning provide predictive capability, while MCDM methods improve prioritisation, transparency, and managerial interpretability. The most common application areas include supplier selection, risk assessment, resilience evaluation, and logistics planning. At the same time, the literature reveals persistent limitations, including reliance on synthetic or simulation-based datasets, limited real-world validation, and challenges in balancing predictive performance with explainability. The paper contributes a structured overview of current research trends and highlights the need for more robust, real-time, and interpretable hybrid frameworks to support resilient and sustainable supply chain decision-making under uncertainty.