A review of research from 2015 to 2025 finds growing use of AI models to predict supply-chain disruptions and support more adaptable responses.
Researchers combined a systematic literature review with an analysis of publication patterns to examine AI applications in supply-chain risk assessment. They reviewed 1,903 articles published between 2015 and 2025 and selected 54 studies using PRISMA guidelines, a standard process for screening research.
The review identified growing use of models such as Random Forest, XGBoost and hybrid systems. It also highlighted explainable AI for more transparent decisions, real-time data and blockchain-based tracking, while identifying China and the United States as major research hubs.
What the review found
The reviewed studies reported that machine-learning models, including Random Forest, XGBoost and hybrid approaches, significantly improve the accuracy and adaptability of supply-chain risk prediction, including in post-pandemic settings. The bibliometric analysis mapped influential authors, institutions and research trends, with China and the United States emerging as leading research hubs.
The review also identified several practical directions: using explainable AI so decision-makers can better understand model outputs, incorporating real-time data, and using blockchain to improve traceability. It emphasized dynamic risk-management strategies, collaboration across disciplines and ongoing evaluation of models.
