AI and Data Centers Consume Energy, Water and Emit Pollution on Par with Nations

A recent report finds that AI systems and data centers consume large amounts of energy and water. Their combined resource use and emissions are comparable to those of many nations. The analysis quantifies the environmental footprint of training and operating AI models. Water consumption stems largely from cooling requirements in large facilities. Pollution includes greenhouse gases released by electricity generation for data centers. The findings raise concerns about the sustainability of rapid AI expansion. Industry groups are urged to adopt greener practices and improve efficiency. Regulators and investors may increasingly scrutinize the climate impact of AI workloads.