Assessing the Environmental Impact of AI-Based Medical Diagnosis Using Grey Relational Analysis

Purpose: This study evaluates the ecological footprint of Artificial Intelligence (AI)-driven healthcare solutions using Grey Relational Analysis (GRA), focusing on energy consumption, carbon emissions, electronic waste (e-waste), and natural resource depletion.Study design/methodology/approach: A multi-criteria decision-making approach integrating the Fuzzy Best Worst Method (FBWM) and the Decision-Making Trial and Evaluation Laboratory (DEMATEL) was employed to assess and prioritize the key environmental factors.Sample and data: Data for FBWM and DEMATEL were systematically collected from a panel of experts. These professionals, with extensive experience in the field, provided critical input on the relative importance of criteria and their interdependencies.Results: The findings highlight that electronic waste, natural resource depletion, and energy consumption are the most significant environmental concerns. At the same time, inefficiencies in AI training and carbon footprint emerge as key causal factors. Addressing these upstream drivers can significantly reduce AI’s overall environmental burden in healthcare.Originality/value: This research provides valuable insights for policymakers and healthcare institutions advocating sustainable AI adoption strategies that balance technological advancement with ecological responsibility.Research limitations/implications: AI development in healthcare requires a shift towards energy-efficient computing, optimized AI training methodologies, and responsible e-waste management. Healthcare institutions, policymakers, and AI developers must collaborate to ensure that medical AI innovations are both technologically transformative and environmentally responsible. Adopting green AI practices—such as hardware recycling, federated learning, and energy-efficient model deployment—is crucial to achieving long-term sustainability in AI-powered healthcare.