By Abbas Nazil
Vast data centers that power artificial intelligence are generating alarming environmental impacts, new research reveals.
Beyond their well-known high energy consumption, these facilities are creating “heat islands,” raising local land temperatures by up to 16 degrees Fahrenheit and affecting the daily lives of over 340 million people globally.
The study, led by Andrea Marinoni, an associate professor with the Earth Observation group at the University of Cambridge, analyzed temperature data collected over the past 20 years and compared it to the locations of large AI “hyperscale” data centers.
These hyperscalers, which house thousands of servers and can span over a million square feet, have rapidly increased in number over the last decade, often outside densely populated urban areas.
By isolating data from seasonal variations, global warming trends, and other influences, researchers determined that surface temperatures rose by an average of 3.6 degrees Fahrenheit near these centers, with extreme cases showing increases of up to 16.4 degrees.
The effects were not limited to immediate surroundings. Temperatures were found to rise up to 6.2 miles away, impacting millions of people in nearby communities.
Regions such as Mexico’s Bajio and Aragon in Spain showed unexplained local temperature increases of approximately 3.6 degrees over the last two decades, coinciding with the establishment of large data centers.
Experts warn the findings are particularly concerning given projections for rapid expansion of AI data centers in the coming years, which could exacerbate heat waves already intensified by global warming.
Deborah Andrews, emeritus professor at London South Bank University, noted this study is among the first to focus specifically on heat generation from data centers, emphasizing the urgency of sustainable planning in the rush to expand AI infrastructure.
Some researchers urge caution, suggesting more studies are needed to verify the magnitude of the reported temperature effects, while noting that overall emissions from powering data centers remain a major climate concern.
Marinoni hopes the research will spark discussion on strategies to mitigate AI’s environmental footprint without hindering technological progress, encouraging policymakers and industry leaders to explore more sustainable pathways for data center development.
The study highlights a growing tension between technological growth and environmental sustainability, underscoring the need for innovative solutions to manage AI infrastructure while protecting communities and ecosystems from rising temperatures.
