Skipping pleasantries like “please” and “thank you” when talking to chatbots could save enough energy to power the annual needs of 760,000 residents in sub-Saharan Africa, highlighting the massive but often hidden environmental toll of artificial intelligence, according to a new UN report.
Released by the Institute for Water, Environment and Health under the United Nations University, the UN’s academic arm, the study published on Wednesday also warned that the true cost of AI extended far beyond carbon emissions to include water consumption, land use and electronic waste.
Data centres, which serve as the physical backbone of AI, consumed an estimated 448 terawatt-hours (TWh) of electricity last year. If data centres were a country, they would rank 11th in the world for energy consumption, roughly on par with France.
Advertisement
Generating this amount of electricity required around 4.5 trillion litres (1.2 trillion gallons) of water – enough to fill 1.8 million Olympic-sized swimming pools. The infrastructure demanded 6,900 sq km (2,700 square miles) of land, which is nearly 4.5 times the size of Greater London, the report said.
This staggering footprint is only expected to balloon as AI usage skyrockets. The report projected AI-related electricity consumption to reach 945 TWh by 2030, accounting for nearly 3 per cent of total global electricity demand.
Advertisement
Furthermore, by 2030, AI infrastructure could generate up to 2.5 million tonnes of e-waste each year – the equivalent of discarding 250 Eiffel Towers annually.