AI Data Centers Are Draining Water in Drought Zones
The AI boom has fuelled a race amongst tech giants to create additional data centers, but there’s a price. A new Bloomberg investigation reveals how AI data centers are sprouting up in drought-prone locations, posing severe water consumption issues. With firms such as Amazon, Microsoft, and Google spearheading the charge, towns already dealing with water scarcity now face an additional challenge.
What’s Happening with AI Data Centers ?
Here’s a snapshot of what’s going on with these AI data centers:
- Over 160 new AI data centers have been built in the US since 2022 — a 70% increase from the previous three years.
- Many of these facilities rely on evaporative cooling, using millions of liters of water daily.
- A 100-megawatt data center can use up to 2 million liters of water per day — the same as about 6,500 homes.
- New projects are targeting drought-hit states like Texas and Arizona, including a massive $100 billion OpenAI facility.
- Globally, AI data centers already consume around 560 billion liters of water per year, and this number could double by 2030.
What This Means for People and the AI Industry ?
The growing number of AI data centers in arid locations is not a local problem; it is a global one. These facilities waste water in locations where clean water is already scarce. While tech companies pledge to be “water positive” by 2030, experts contend that offsetting water elsewhere does not address the urgent local crisis.
This broadens the environmental impact of AI for both industry and users. It’s no longer only about energy consumption; water is now a key resource under threat from the never-ending need for AI services and infrastructure.
Our Thoughts !
Honestly, this is one of those things that makes you think. While we are all thrilled about the potential of AI, it is clear that AI data centers have hidden costs. Even with water-positive vows, the reality that digital behemoths are building these in already water-stressed areas feels like a missed opportunity.
We believe the industry must immediately shift toward smarter, more sustainable cooling methods, such as Microsoft’s closed-loop technology, and reconsider where these data centers are located. As AI advances, it should not come at the expense of critical resources such as water.
We’ll be keeping a careful eye on whether innovation catches up with responsibility in this situation. Because, at the end of the day, AI should improve the world rather than make it thirstier.