There has been a lot of talk across the internet about the water usage of AI Data Centers and while they don't consume a bottle of water every couple queries, the waste is real. AI data centers consume up to 5 million gallons of water per day and will keep growing with more demand on AI(Yanez-Barnuevo). The main problem is that when data centers use the water for cooling approximately 80% of it evaporates (Li et al. 3) and with companies like Meta having 48% of their water intake coming from local sources(Bush), data centers are putting strain on communities. This level of water consumption is especially problematic because freshwater is a limited resource. Many data centers are located in regions that already face water scarcity or drought conditions. As a result, these facilities compete directly with local communities for access to clean drinking water (Girolimon).
In addition to water consumption, data centers also produce wastewater, which can place additional strain on local water treatment systems. If not properly managed, this can lead to environmental contamination and further reduce the availability of safe water.
One of the main reasons this problem persists is the lack of strict regulations governing water usage by data centers. In many areas, companies are allowed to draw water from public supplies or underground aquifers without significant restrictions (EESI). There are often no requirements for transparency, meaning companies are not always required to report how much water they use.
Governments may also provide incentives to attract data centers without fully considering the long-term impact on water resources (Gorey). This creates a situation where economic development is prioritized over environmental sustainability and resource conservation.
The communities most affected by water usage from data centers are often those already facing limited access to water. Rural areas and regions prone to drought are particularly vulnerable. When large amounts of water (up to 50,000 households worth per data center) are diverted to support data centers, local residents may experience shortages or increased costs for water services (Girolimon).
Low-income communities are again disproportionately impacted. These populations may lack the infrastructure or political influence to advocate for fair water distribution. As a result, they bear the burden of reduced access to a basic necessity while corporations benefit from the resources (Bush).
Below is a in person example of the affects these data centers can have on access to water and water quality in Georgia. The story follows a woman named Beverly Morris and how she came to this part of rural Georgia to retire but an data center has moved in and she no longer feels safe drinking the water. We may not actively think about the consequences of technological advancement but data centers are actively moving into lower income rural communities and are damaging the quality of life of those around it and as long as we allow data centers to be unregulated they will continue to do so.
Addressing the power problem of AI data centers requires both policy change and public awareness. As citizens, one of the most effective actions is to support policies that promote renewable energy, such as wind and solar power. Voting for leaders who prioritize sustainability and holding companies accountable for their environmental impact are also important steps.
Consumers can also influence change by supporting companies that are committed to reducing their carbon footprint. Raising awareness about the hidden energy costs of AI technologies can encourage more responsible usage and innovation. Finally, advocating for stricter regulations on energy consumption and emissions can help ensure that technological progress does not come at the expense of the environment (UNEP).
I added a petition below so if you feel passionate about the cause after reading this and want to take a step towards action feel free to sign.