In recent years, artificial intelligence has advanced rapidly, transforming how we live and work. Tools like ChatGPT have become household names, offering instant answers, creative support, and business insights at our fingertips.
But behind the convenience is a lesser-known environmental cost: water.
Yes — water.
Every time someone uses ChatGPT, a small but meaningful amount of water is consumed to keep the infrastructure running. AI models like ChatGPT are hosted in large data centers, many of which rely on advanced cooling systems to prevent servers from overheating.
These systems often use evaporative cooling — a process that consumes significant amounts of water to maintain safe operating temperatures.
So how much water does one interaction use?
According to a 2023 study from researchers at the University of California, ChatGPT may use around 500ml of water for every 20 to 50 queries.
That’s roughly equivalent to a standard bottle of water for a single extended interaction.
Multiply that across millions of users globally, and the scale becomes clear.
The bigger picture: training AI models
At an individual level, the impact may seem small.
But the training phase of large AI models tells a different story.
Training models like GPT-3 can require thousands of high-performance GPUs running continuously for weeks or months. During this process, large volumes of water are used to cool the hardware.
Estimates suggest that training GPT-3 alone consumed over 700,000 liters of fresh water — primarily for cooling during intensive computation.
Why water?
Cooling is essential to maintain both performance and hardware lifespan in data centers.
While some facilities use air-based cooling, water-based systems are often more energy-efficient and cost-effective — but they come with a significant environmental trade-off.
In regions where water is scarce, or where freshwater supplies are already under pressure, this demand can place additional strain on local ecosystems.
What does this mean for sustainability?
As businesses and individuals become more reliant on AI tools, it’s important to understand their full environmental footprint — not just in terms of electricity and carbon emissions, but also water usage.
AI companies are actively working to improve efficiency and adopt more sustainable data center practices.
But user behavior also matters.
More intentional use — especially at scale — can help manage demand and support more sustainable innovation.
A global perspective
As climate change intensifies water scarcity in many parts of the world, the hidden water usage behind digital tools becomes more than a technical detail.
It becomes a global environmental concern.
As we continue to adopt AI, the challenge is to balance innovation with responsibility — ensuring that the unseen costs of progress don’t outweigh its benefits.