The rapid expansion of artificial intelligence will result in significant environmental consequences extending far beyond electricity consumption and greenhouse gas emissions. A new study by the United Nations University Institute for Water, Environment and Health (UNU-INWEH) examines the broader environmental impact of AI, including the water and land use associated with the electricity needed to power AI infrastructure. The report, "Environmental Cost of AI's Energy Use: Carbon, Water and Land Footprints", also highlights the material footprint of AI, from its reliance on critical raw materials and specialised hardware to the growing challenge of managing waste electrical and electronic equipment (WEEE).
AI is not merely software but depends on a vast physical infrastructure and supply chain, the authors point out. This includes chips and data centres, power generation facilities, cooling systems, water withdrawals, land use and critical raw materials, as well as the electronic waste generated when equipment reaches the end of its life. According to the authors, AI infrastructure could generate up to 2.5 million tonnes of e-waste a year by 2030, equivalent to roughly 250 Eiffel Towers....
