AI is Using Our Drinking Water: The Hidden Thirst of Data Centres

AI is using our drinking water through data centres cooling.

AI is booming. From ChatGPT to self-driving cars, artificial intelligence is powering the future.
But behind this digital revolution, there’s a quiet, invisible cost: our drinking water.

Introduction & Background

When we think about AI, we usually imagine algorithms, servers, and endless lines of code. What we don’t think about is water. Yet, to keep data centre running—the massive facilities that power AI models—millions of liters of fresh water are consumed every single day. This water isn’t just any water. It’s the same clean, drinkable water that could otherwise serve households, agriculture, and ecosystems. So why exactly does AI need our water?

Cooling Data Centers: Why AI is Thirsty

Data centers are essentially huge warehouses packed with thousands of computers. These computers run nonstop, generating enormous amounts of heat.
If they overheat, they break down. To prevent this, companies use cooling systems—and that’s where water comes in.

Evaporative cooling is the most common method: fresh water is circulated to absorb heat, then released as vapor.
Running advanced AI models like GPT-4 requires significantly more energy than traditional computing, which in turn demands more cooling.
A single AI query, like asking ChatGPT a question, indirectly consumes a fraction of a liter of water when scaled across millions of users.

In other words, every time you chat with AI, somewhere in the world, water is being used to keep the machines alive.

How Much Water Are We Talking About?

The numbers are staggering.

Research from the University of California, Riverside, estimates that training GPT-3 used around 700,000 litres of clean water—enough to produce hundreds of cars or grow tons of wheat.
Microsoft reported that its global water consumption jumped 34% in a single year, largely due to AI expansion.

This hidden water footprint is becoming an urgent concern, especially in regions already struggling with droughts and water shortages.

Sam Altman’s Statement on the Issue

Even Sam Altman, CEO of OpenAI, has acknowledged this problem. He admitted that AI’s environmental impact—particularly its water and energy use—is one of the biggest challenges the industry faces.

Altman has suggested that while AI could solve humanity’s hardest problems, including climate change, the industry must first address its own sustainability crisis.
In simple words: AI can’t save the planet if it’s drinking it dry.

The Bigger Picture: Technology vs. Sustainability

This raises tough questions:

Should tech companies be allowed to consume fresh water at such massive scales?
Can AI’s growth be balanced with sustainable resource management?
Will innovations in cooling—like using recycled water, seawater, or advanced cooling tech—reduce this hidden cost?

Some companies are experimenting with alternatives, like immersion cooling (submerging servers in special liquids) or using non-drinkable water sources. But large-scale adoption is still a long way off.

Current Progress & Hope for the Future

The good news is that awareness is growing. Environmental researchers, governments, and even AI companies are beginning to address the issue.

Google and Microsoft have pledged to become “water positive” by 2030—meaning they’ll replenish more water than they consume.
Data centers in cooler climates are reducing dependence on freshwater by using outside air cooling.

But whether these solutions can keep up with the explosive growth of AI is still uncertain.

Conclusion

AI promises to change the world—but at what cost?
The reality is that AI is using our drinking water, and unless urgent steps are taken, this could worsen water scarcity in already vulnerable regions.

As Sam Altman admitted, the tech industry must reckon with this truth: innovation cannot come at the expense of survival.
Because at the end of the day, no amount of artificial intelligence can replace a glass of clean, fresh water.