Machine Learning Exposes Hidden Conflict Risks in Global Mineral Supply (opens in new tab)

Researchers developed a machine learning framework to map ESG-driven conflict risks across energy transition mineral mining projects, using 112,766 conflict events and 16 environmental, social, and governance indicators. The study found that environmental factors drove most modeled risk, with tungsten showing the highest overall risk, platinum the lowest, and lithium most affected by water-stress-related environmental risk.