Report Reveals Rising Environmental Impact of Data Centers Driven by AI
A United Nations University report highlights the environmental footprint of global data centers, predicting that their energy and water usage will double in the next four years due to increased reliance on artificial intelligence.
The environmental footprint of global data centers now rivals that of some of the world's largest nations, according to a recent report from the United Nations University. In 2022, data centers consumed 448 trillion watt-hours of electricity, translating to approximately 208 million tons of carbon dioxide emissions, comparable to those of Argentina. The report suggests that by 2030, the electricity use of data centers could reach 935 trillion watt-hours, accounting for nearly 3% of the world’s total electricity consumption and generating about 440 million tons of carbon dioxide.
The report emphasizes that the surge in energy consumption is largely driven by the growing demand for artificial intelligence. Currently, about 20% of the energy utilized by data centers is attributable to AI, a figure expected to rise to 40% by 2030. Kaveh Madani, a co-author of the study, noted, "If you look at these numbers, we’re seeing scales comparable to nations. The demand is enormous."
Experts have praised the report for its comprehensive approach to environmental impacts, with Fengqi You from Cornell University stating that it effectively integrates issues of carbon emissions, water usage, and environmental justice. Jean Su, director of the Energy Justice Program at the Center for Biological Diversity, highlighted the report as a pivotal piece of research shedding light on the ecological consequences of AI.
Caleb Max, president of the National Artificial Intelligence Association, emphasized AI's potential benefits, arguing that it brings improvements to daily life and productivity. Concurrently, Josh Levi, president of the Data Center Coalition, reiterated the industry's commitment to addressing environmental concerns responsibly and transparently.
Madani also outlined ways users can mitigate energy consumption by modifying their AI queries. He explained that simplifying requests by 30% could lead to a 25% reduction in energy use, which would amount to the annual electricity consumption of approximately 700,000 people in Africa.
Despite advancements in efficiency, Madani pointed out a paradox: as technology becomes more efficient, total energy use can still increase due to heightened utilization. Researchers noted transparency from companies regarding their energy and resource consumption remains a significant challenge.
