# Environmental Impact of AI

> The power, water and stuff it takes to train and run AI.

**Field:** Artificial Intelligence  
**Category:** Ethics & Regulation  
**Also known as:** AI Energy Use, AI Carbon Footprint, AI Water Use

## Think of it like

It is like leaving a big oven on all day. Each meal costs power, and it adds up.

## Definition

The environmental impact of AI is the power, water and materials it takes to build and run it. Models are trained and served in data centres full of chips. These use a lot of electricity, and many use water to keep the chips cool. Making the chips takes metals and energy too. Data centres used about 1.5% of the world's electricity in 2024, and that is set to more than double by 2030. AI is the biggest reason for the growth.

## Example

Emma's company picks a small model for its chatbot, not the biggest one. Each answer uses far less computing, so the power used drops a lot.

## Why it matters

Every answer has a cost in power. At billions of questions a day, small choices add up.

## Common mistake

Counting only the training run. Each answer uses power too. For a busy model, that adds up fast.

## Three principles

- **Power to run.** Chips use a lot of electricity.
- **Water to cool.** Many data centres use water.
- **Smaller is lighter.** Use the smallest model that works.

## Related terms

- [Small Language Model (SLM)](https://learn.benibix.com/term/ai/small-language-model)
- [Parameter Count](https://learn.benibix.com/term/ai/parameter-count)
- [Inference (AI)](https://learn.benibix.com/term/ai/inference)
- [Training (AI)](https://learn.benibix.com/term/ai/training)
- [Responsible AI](https://learn.benibix.com/term/ai/responsible-ai)
- [Graphics Processing Unit (GPU)](https://learn.benibix.com/term/infrastructure/graphics-processing-unit)

![Four panels. Jobs: a row of four tasks, one handed to AI first. Power and water: data centre electricity for the world, 415 TWh in 2024 and about 945 TWh forecast for 2030. Watching: one camera picking a single face out of a crowd. Weapons: two boxes, a person decides or the machine decides.](https://learn.benibix.com/diagrams/ai4-ai-society.svg)

---

Source: https://learn.benibix.com/term/ai/environmental-impact-of-ai
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-18
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
