Power plants draw heavy water use to meet AI electricity demand

Water strain from AI extends beyond data centers, as plants producing the electricity for it also consume large volumes. Concern over the carbon footprint of data centers is also growing in climate debates.

By Standard Topic. AI-generated from the reporting listed under Coverage. Published October 15, 2026. Updated October 15, 2026.

This story is no longer being updated. It reflects reporting up to October 15, 2026.

What happened

Power plants are consuming large volumes of water to generate the electricity that runs artificial intelligence, including its more efficient cooling systems. Artificial intelligence has also become a lightning rod in climate debates over the carbon footprint of the data centers that power it. A reader asked whether running large language models on personal computers could spare shared data centers, and the answer was that small computers cannot always fix the problem.

Why it matters

The water and carbon demands tied to AI raise concerns about the technology's environmental cost. Shifting models to personal computers is not a guaranteed remedy.

Timeline

Oct 14, 2:32 PM ET: Power plants' water consumption for AI electricity is highlighted as a growing strain.
Oct 14, 9:48 PM ET: AI's carbon footprint from data centers is discussed as a possible climate tool.
Oct 15, 6:00 AM ET: Storing large language models on personal computers is raised as a way to save resources, with the answer that small computers cannot always fix the issue.

Coverage

EarthTalk: Small computers can’t always fix LLM hunger issue (hometownsource.com) October 15, 2026
Behind AI's carbon footprint, a potential climate tool (Yahoo News) October 14, 2026
AI's water strain includes surging demand from electricity use (Los Angeles Times) October 14, 2026

Power plants draw heavy water use to meet AI electricity demand

What happened

Why it matters

Timeline

Coverage

More from artificial-intelligence