AI Data Centers in Georgia:

Effects on Water and Living Space 

By: Mark Shehata 

 

Ai data centers are one of the largest current topics in the United States and the world in general. The focus on data centers occurs because of the sudden need to process digital services for increased use of AI services. This is expected to grow in the next few years. For example, in 2023 US data centers used 176 terawatt hours of electricity and it is expected to grow to 325-580 terawatt hours of electricity in 2028 (George, 2026).

 

The massive use of power ends up creating several environmental factors that affect people's health. Such as creating heat centers in urban and rural areas, air pollution, water scarcity and pollution, noise pollution. Lastly, there is also an environmental justice element to this issue due to these data centers being built near populations who cannot resist the goals of corporations who want land to build them such as in Georgia.  

 

One of the lesser known effects of data centers is the creation of local climate zones. The energy consumption, generation of air pollutants, and usage of fossil fuels create zones where the temperature increases by 2 celsius on average (Maroni et.al, 2026). This heat island effect contributes to climate change. Air pollution is another side effect, diesel generators in data centers can emit up to 12 metric tons of nitrous oxide annually during testing as seen in Texas (George, 2026). This leads to increased risk for respiratory disease in these areas. Noise pollution occurs when generator testing and cooling systems create a low frequency hum that interrupts daily living and sleep. Environmental noise literature shows that the stress and sleep loss from noise pollution contributes to increased heart disease and premature death (George, 2026). 

 

Water usage for cooling is also a big issue with data centers. A medium size data center can consume around 110 million gallons of water per year which is around the same amount for 1,000 US households (Yanez-Barnuevo, 2025). While most water used ends up evaporating, the potential discharge from this cooling process returns warm water back to the environment that increases bacteria growth in waterways. Water returned from data centers can end up having higher salinity, heavy metals, and other contaminants (George, 2026). While studies on water scarcity issues exist, more research is needed on the health effects of water.

 

Finally, the issue of environmental justice is important in regards to the political dimensions of data centers. Data centers are often built in areas that already have environmental burdens and increase air and noise pollution (George, 2026). This can also be seen in Georgia where Georgia power is attempting to use eminent domain in order to build a transmission line where houses dwell (Henry & Luibrand, 2026). Data centers can be built without these negative issues. There are a wide variety of methods that can be used, such as noise regulation, land use planning, water governance, health impact assessment and monitoring, and environmental justice safeguards. By implementing safeguards negative health outcomes can be prevented for populations near data centers. 

Works Cited 

​

George, B. (2026). Health Consequences of Large Data Centers: Air Pollution, Noise, Water Use, and Environmental Justice. No. Cnwyb_V1.DOI:10.31235/osf.io/cnwyb_v1 

Henry, S., & Luibrand, S. (2026, July 15). Georgia family says they’re forced to sell home to help power AI data centers: “it’s theft.” CBS News. https://www.cbsnews.com/news/georgia-power-ai-data-centers-eminent-domain/   

​

Marinoni, A., Cambria, E., Lin, W., Mura, M. D., Chanussot, J., Ragusa, E., ... & Horton, B. (2026). The data heat island effect: quantifying the impact of AI data centers in a warming world. arXiv preprint arXiv:2603.20897. https://doi.org/10.48550/arXiv.2603.20897 

Yanez-Barnuevo, M. (2025, June 25). Data Centers and Water Consumption. Environmental and Energy Study Institute (EESI). https://www.eesi.org/articles/view/data-centers-and-water-consumption