ON
← Back to feed
WorldEnvironment3 days ago

AI could be trapped in a 'Carbon Valley' unless action is taken soon

A new study published in the journal Communications Earth & Environment warns that rapid AI development could lead to significant short-term carbon emissions, creating what researchers call a 'Carbon Valley.' The research, led by Yassine Charabi of Kuwait University, uses simulations based on global energy forecasts, data center growth, and chip manufacturing to show that AI's environmental benefits may not offset its initial carbon costs for many years.

AI is growing fast, and keeping up means building more data centers, manufacturing advanced chips and powering the tech behind it. All of that comes with a carbon cost. AI advocates claim that in the long run, AI will save energy and cut carbon emissions across global industries.

However, a new study published in the journal Communications Earth & Environment points out a flaw in this argument. By the time those long-term savings come, we would already have accumulated a substantial cumulative carbon debt, according to its calculations.

Yassine Charabi, a geographer at Kuwait University, built a mathematical simulation to map out how AI expansion affects the planet over time. He plugged in data such as global energy forecasts, data center growth rates and hardware replacement schedules. The model then factored in everyday electricity use, emissions from grid electricity and the carbon generated during chip manufacturing.

Trapped in the Carbon Valley

Charabi ran the simulation 10,000 times, and the results revealed a long stretch of negative impact he calls the Carbon Valley . This is when building and powering AI produces more carbon pollution than the technology can save.

Under the fastest AI growth scenario, the model shows that AI will stay trapped in this valley for nearly a decade, with AI-enabled savings failing to catch up until late 2031. By that time, the world will have already accumulated a peak cumulative carbon debt of around 2.85 gigatons of carbon dioxide.

But it doesn't have to be this way.

Charabi's model shows that to shrink the Carbon Valley, we will need to accelerate AI deployment in green industrial processes , because each year of delay adds 0.45 gigatons of CO 2 .

The study finds that the climate benefits depend on how much AI is actually tied to emissions-reducing uses. Also, location matters, since cooler climates reduce the amount of cooling-related energy used compared with hotter regions.

The cost of hesitating

Relying on the technology simply to become more efficient will not solve the problem, as Charabi notes in his paper.

"Efficiency improvements alone do not ensure absolute decoupling between AI expansion and electricity demand under rapid deployment pathways."

If these approaches do not come soon, clean energy fixes may arrive too late to reverse the initial damage, he warns. "Under the accelerated AI data-center deployment scenario analyzed here, subsequent reductions in annual emissions alter future trajectories but do not erase earlier additions to cumulative CO₂ within the cumulative-emissions accounting framework adopted in this study."

Written for you by our author Paul Arnold , edited by Gaby Clark , and fact-checked and reviewed by Robert Egan —this article is the result of careful human work. We rely on readers like you to keep independent science journalism alive.

If this reporting matters to you, please consider a donation (especially monthly). You'll get an ad-free account as a thank-you.

Publication details

Yassine Charabi, Rapid artificial intelligence deployment increases near-term pressure on global carbon budgets, Communications Earth & Environment (2026). DOI: 10.1038/s43247-026-03746-y

Who's behind this story?

Paul Arnold

BSc Biology from University of London. BBC documentary producer with world travel experience. Freelances from southern Spain.

Full profile →

Gaby Clark

MA in English, copy editor since 2021 with experience in higher education and health content. Dedicated to trustworthy science news.

Full profile →

Robert Egan

Bachelor's in mathematical biology, Master's in creative writing. Well-traveled with unique perspectives on science and language.

Full profile →

© 2026 Science X Network

Citation :

AI could be trapped in a 'Carbon Valley' unless action is taken soon (2026, June 17)

retrieved 17 June 2026

from https://phys.org/news/2026-06-ai-carbon-valley-action.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no

part may be reproduced without the written permission. The content is provided for information purposes only.

Read the full article at Phys.org
Source document: Analysis of 97 global markets

2 reports

QuartzIndependentCenter3 days ago
AI data centers face threats from flooding, wildfires, and extreme heat, study warns

An analysis of 97 global markets found that chronic stress from heat and drought threatens more than half of all data centers worldwide.

Bias read (Center): The article presents a factual warning about environmental risks to data centers without overtly favoring any political perspective. It focuses on technical and environmental factors rather than policy or ideological debates.

Official sources cited

  • study Analysis of 97 global markets
Phys.orgIndependentCenter4 days ago
AI could be trapped in a 'Carbon Valley' unless action is taken soon

A new study published in the journal Communications Earth & Environment warns that rapid AI development could lead to significant short-term carbon emissions, creating what researchers call a 'Carbon Valley.' The research, led by Yassine Charabi of Kuwait University, uses simulations based on global energy forecasts, data center growth, and chip manufacturing to show that AI's environmental benefits may not offset its initial carbon costs for many years.

Bias read (Center): The article presents findings from a scientific study without overtly favoring any political stance. It reports on the environmental implications of AI development using technical details and does not include biased language, one-sided sourcing, or editorial commentary that would indicate a leaning.

Official sources cited

Go to the primary sources (2)

The official sources this coverage is built on. Read them directly to bypass framing.