Environmental impact of AI-driven data center growth in the U.S.
The rapid growth of AI chatbots is fueling an explosive demand for data centers, particularly in areas like Loudoun County, Virginia. This expansion is exacerbating environmental concerns by prolonging the life of fossil fuel power plants, spurring new fossil fuel infrastructure, and posing risks to water resources across the U.S. Experts highlight the substantial energy and water consumption of AI training and deployment, calling for increased transparency and regulation in the industry.
Sarina Virmani, a high school student from Loudoun County, Virginia, a region known as Data Center Alley, has published research on the environmental impact of data centers and advocates for greater industry regulation and transparency. She emphasizes that the perceived invisibility of AI belies its physical manifestation in massive, energy-intensive data centers.
The proliferation of AI chatbots like OpenAI's ChatGPT, Anthropic's Claude, and Google's Gemini is driving a surge in data center demand, which is linked to several environmental issues. These include extending the operational life of aging oil, gas, and coal infrastructure, prompting the construction of new fossil fuel facilities, and raising concerns about water resource depletion.
Experts, such as AI scholar Sajjad Moazen, note that training a single large language model like ChatGPT-3 can consume as much electricity as over 1,000 U.S. households annually. Quentin Good of the Frontier Group, who calls these extended fossil fuel plants "zombie power plants," states that at least 15 such plant retirements have been postponed due to increased energy demand, leading to significant climate-changing pollution.
Data center water usage is also a concern, particularly in water-stressed regions. While not a crisis in Virginia, according to Good, arid areas like Colorado face greater risks, especially during hot, dry months when water demand peaks. Furthermore, the discharge of hot, chemically altered water from data centers can negatively impact local waterways and wildlife. The article concludes by noting the bipartisan interest in AI industry regulations and the potential for local communities to influence data center development.