AI infrastructure surge brings unexpected climate consequences

The rapid expansion of artificial intelligence infrastructure is bringing new environmental concerns beyond its already significant energy consumption.

New research has revealed that large-scale data centers—especially AI “hyperscalers”—are contributing to localized warming effects known as “heat islands,” as highlighted by a Fortune article, increasing surrounding land temperatures and affecting hundreds of millions of people worldwide.

These hyperscale facilities, which house thousands of servers and can span over a million square feet, require immense amounts of energy not only for computation but also for cooling systems that prevent overheating. The study, led by Andrea Marinoni, an associate professor with the Earth Observation group at the University of Cambridge, highlights how this energy use translates into heat released into the surrounding environment.

Marinoni noted that there are still major gaps in understanding the broader environmental impacts of data centers, particularly as their numbers grow rapidly across the globe.

To investigate this issue, researchers analyzed two decades of temperature data collected from remote sensors and compared it with the locations of over 6,000 data centers. These sites were specifically chosen in less densely populated areas to reduce interference from other heat sources such as industrial activity or residential heating.

The team also accounted for seasonal variations, long-term global warming trends, and other environmental factors to isolate the effect of the data centers themselves.

Their findings revealed that surface temperatures increased by an average of 2 degrees Celsius after a data center became operational. In more extreme cases, nearby temperatures rose by as much as 9.1 degrees Celsius. These patterns were observed consistently across different regions of the world. For example, Mexico’s Bajio region, now a major data center hub, experienced unexplained temperature increases of around 2 degrees over the past 20 years. Similarly, Aragon in Spain showed comparable warming trends that were not seen in neighbouring provinces.

The study also found that the impact extends far beyond the immediate vicinity of the facilities. Temperature increases were detected up to 10 kilometres away, collectively affecting more than 340 million people. This widespread influence raises concerns about how expanding AI infrastructure could intensify existing climate challenges, particularly as global temperatures continue to rise.

The planned scale up of data centers “could have dramatic impacts on society” in terms of the environment, people’s welfare and the economy, Marinoni said.

By Nazrin Sadigova