AI's Environmental Impact: Emissions Equivalent to Millions of Cars

A study reveals that the growth of AI in the US may result in carbon emissions similar to adding millions of cars on roads, with substantial impacts on water consumption. Researchers suggest sustainable strategies to reduce AI's environmental footprints, highlighting data center locations and efficient operations.

In a groundbreaking study published in Nature Sustainability, researchers have raised alarms about the growing environmental impact of artificial intelligence in the United States. The study estimates that the rapid expansion of AI could result in 24 to 44 million metric tonnes of carbon dioxide emissions yearly, comparable to adding between five and ten million cars on American roads.

Besides carbon emissions, the study also projects that AI's increasing usage may consume 731 to 1,125 million cubic meters of water annually, equivalent to the annual household water consumption of 6 to 10 million Americans. Such findings are critical, given AI's prominence in transforming various sectors, from finance to manufacturing.

Lead researcher Fengqi You from Cornell University emphasized the urgency of adopting sustainable practices, including smart siting of data centers and improved cooling efficiencies, to dramatically slash AI's environmental impacts. Even with the best practices, however, existing clean energy constraints may limit the potential for reductions in emissions and water usage.

(With inputs from agencies.)