Whether you like it or not, Artificial Intelligence (AI) has become a major part of modern life, from chatbots to algorithms to self-driving cars. AI is revolutionizing industries, optimizing logistics, automating customer service, and even aiding scientific research. However, the convenience that AI offers comes with an extensive environmental cost that often goes unnoticed.
Every time you talk to an AI chatbot like ChatGPT, your request is processed through large-scale data centers that rely on massive computing power to generate responses. These facilities require extensive cooling systems to prevent overheating, and a significant portion of these systems rely on fresh, clean water—often consuming local resources at an alarming rate.
To put this into perspective, generating a short response from AI uses about 500 milliliters of fresh water— about as much as a standard water bottle. While this may seem like a small amount, the cumulative impact is staggering when considering the millions of AI interactions that occur daily across the globe. Data centers, facilities used to house computer systems and data, are the backbone of AI operations.
“Google’s data centers in the U.S. alone consumed an estimated 12.7 billion liters of fresh water in 2021 to keep their servers cool,” the University of California, Riverside found in this study by Bourns College of Engineering.
This number is only a fraction of the total global water consumption from AI-related computing activities, especially in 2024. Training a large AI model can require using millions of liters of water. This process involves running high-powered servers for weeks or months at a time, drastically increasing energy and water consumption. Beyond simple AI interactions, training large AI models significantly amplifies water consumption. This process involves running high-powered servers for weeks or even months, leading to the use of millions of liters of water. The sheer demand for water resources is expected to rise as AI becomes further integrated into various aspects of life, potentially straining local water supplies and contributing to worsening water scarcity.
In regions already facing droughts and water shortages, AI-driven data centers can exacerbate existing problems. Communities that rely on local water supplies for agriculture, drinking water, and industry may find themselves competing with massive tech companies for an increasingly precious resource.
Beyond just water usage, AI also contributes to climate change through its high energy consumption and carbon emissions. AI models rely on specialized hardware like tensor processing units (TPUs) and graphics processing units (GPUs) to handle complex computations. These processors require immense amounts of power, leading to a significant carbon footprint.
In 2019, researchers from the University of Massachusetts, Amherst, found that training a single large AI model could produce as much carbon emissions as five cars over their entire lifetimes. These staggering results highlight the immense amount of energy AI truly consumes. The widespread deployment of AI has only increased this energy burden.
The source of this energy also calls for concern. Many AI data centers still rely on fossil fuel-generated electricity, contributing to greenhouse gas emissions and accelerating climate change. Although some tech companies are moving towards renewable energy, the process is slow.
Developing more efficient cooling systems that rely on alternative methods, such as liquid or air-based cooling, or renewable energy is extremely important. On the user end, minimizing unnecessary AI usage can also help to reduce overall consumption.
The rapid growth of AI means that addressing its effect on the environment is more important than ever. Without strategic intervention, AI usage could worsen water scarcity, increase carbon emissions, and place additional stress on global energy resources. However, by investing in energy-efficient models and responsible AI development, we can work towards a future where AI innovation does not come at the expense of the planet. Governments, corporations, and researchers must all work together to establish green AI policies, improve resource efficiency, and ensure accountability for the environmental impact of AI systems. The challenge is clear: if AI is to be a force for progress, it must also be a force for sustainability.
