Selecting Next-Generation AI Technologies for Sustainable Pharmaceutical Supply Chains: An Integrated LOPCOW–RAFSI Approach

The purpose of this study is to evaluate and prioritize next-generation artificial intelligence (AI) technologies for sustainable pharmaceutical supply chains by considering multiple technological, operational, sustainability, and regulatory criteria. Within the methodology of the study, an integrated LOPCOW–RAFSI multi-criteria decision-making framework was employed. Fifteen evaluation criteria were identified through a comprehensive literature review and reviewed for domain relevance and applicability, and six AI technology alternatives—Generative AI, Agentic AI, AI-powered Digital Twin, Federated Learning, Explainable AI, and Reinforcement Learning—were evaluated. The LOPCOW method was used to determine objective criterion weights, while the RAFSI method was applied to rank the alternatives. The findings indicate that AI-powered Digital Twin (A3) achieved the highest RAFSI score (0.666060), followed by Agentic AI (A2) (0.547032) and Explainable AI (A5) (0.541363), resulting in the overall ranking A3 > A2 > A5 > A6 > A4 > A1. The sensitivity analysis further showed that A3 retained the first position across all examined scenarios (S0–S6), although limited rank reversals occurred between A2 and A5. As a contribution to the literature, this study integrates objective weighting and robust ranking within a unified decision-support framework for evaluating emerging AI technologies and provides decision-makers with a systematic and transparent basis for prioritizing AI investments in sustainable pharmaceutical supply chains.