The global transition to renewable energy hinges on critical minerals like lithium, cobalt, platinum, antimony, and tungsten, but their extraction and supply chains are imperiled by environmental, social, and governance risks that undermine stability and sustainability. Here, we introduce a novel framework for environmental, social, and governance risks evaluation, fusing a multi-dimensional indicator system with advanced machine learning models. Drawing on conflict events alongside diverse socio-economic and environmental indicators, we predict environmental, social, and governance risks at a global scale. Results uncover marked spatial heterogeneity, with elevated risks in the Middle East, North Africa, and South Asia associated with governance weaknesses and environmental pressures. Tungsten exhibits the highest overall environmental, social, and governance risks, while platinum the lowest. Our findings underscore the need for targeted, dimension-specific mitigation strategies and enhanced international cooperation to bolster sustainable governance and secure critical mineral supplies for a resilient energy transition. Tungsten shows highest and platinum lowest environmental social and governance risk, with hotspots in Middle East North Africa and South Asia, predicted using multi-dimensional indicators and machine learning from 112766 conflict events.
Zhu et al. (Thu,) studied this question.