Energy Transition, AI, and Geopolitics

The Invisible Cost of AI: Why Data Centers are Straining America’s Water Supply

By Ken Silverstein • March 26, 2026 • Filed in: Uncategorized

massive data center exterior with cooling towers and water infrastructure, AI generated

The AI revolution is colliding with a finite physical reality: water.

In my latest piece, I look at the ‘unplanned’ water problem facing the AI boom. With companies like Microsoft and Google racing to scale, the strain on municipal water supplies is becoming a primary regulatory and operational hurdle.

The Data-Water Nexus: A New Infrastructure Priority

How should we prioritize industrial cooling vs. residential supply in water-stressed regions? It’s a conversation that needs to happen at the planning level today.

The strategic vulnerability here isn’t just about environmental footprints; it’s about the physical limits of the digital frontier. While we have spent decades building regional power grids, our water infrastructure remains hyper-local. When a hyperscale data center requires 300,000 gallons of water per day, it isn’t just a tech project—it becomes a competitor for a town’s most vital resource. As I noted in my coverage for Forbes, the solution isn’t fewer data centers, but a shift toward regional planning that treats water as a primary constraint rather than a utility afterthought.

This mirrors the volatility we are seeing in global natural gas markets, where supply disruptions are increasingly tied to infrastructure vulnerabilities rather than to fuel availability alone.

The Forbes story

Comments are closed.

 

« | Home | »

  • Recipient of the ASBPE Gold Award for Outstanding Web Commentary and the MIN Online “Most Intriguing People in Media” honor. Senior Contributor at Forbes with nearly 30 years of energy and climate reporting experience.