Truace
Subscribe
52/100

Moderate evidence

Cite
More
Export CSVExport JSONEmailLinkedInX

Gulf AI data centers: model finds emissions rise by 2035 and AI offsets fall short

Study title: Net sustainability assessment of AI-driven data centers in the GCC: an MCDA approach framework

Evidence type
Benchmark or lab evaluation
Sources
1 sourcePeer-reviewed
Published
3 min read
Truvace record
TRV-2026-2605cite or check it

Study title: Net sustainability assessment of AI-driven data centers in the GCC: an MCDA approach framework

A multi-criteria scenario model of AI-driven data center growth in the six Gulf Cooperation Council countries finds that, under a business-as-usual high-growth path, modelled emissions increase substantially by 2035. AI sustainability applications in buildings, industry and utilities offset only part of that footprint, even assuming optimistic adoption.

Read the study abstract

The Gulf Cooperation Council (GCC) is witnessing a rapid growth in artificial intelligence (AI)-enabled data center development, thanks to national digital transformation strategies and the region’s strategic location bridging global markets. Yet, no evaluation framework holistically compares the direct impacts of data center operations on the environment with the dispersed sustainability impacts of AI applications across major business sectors. To address this, this study proposes and applies a Multi-Criteria D…

Net sustainability assessment of AI-driven data centers in the GCC: an MCDA approach framework
Data center cooling tower by Rsparks3. CC0 · http://creativecommons.org/publicdomain/zero/1.0/deed.en

51Gain side

How the two sides scoreClose call

54Problem side

The gain

A scenario model of GCC data centers suggests AI applications in buildings, industry and utilities could offset part of the footprint, and net-positive outcomes appear only when renewable procurement, best-practice cooling and verified AI benefits combine.

The problem

A scenario model of GCC data centers suggests that under business-as-usual high growth, emissions rise substantially by 2035, with AI sustainability applications offering only partial offsets even under optimistic adoption assumptions.

A fixed formula sets both scores from the evidence and its sources. Reader votes don’t move them.

What the scores mean

G is the gain side, out of 100. It goes up when the gain is specific, when it measurably helped people and when the sources behind it are strong.

P is the problem side, out of 100. It goes up when the problem is specific, when it measurably hurt people and when the sources behind it are strong.

In brief

The authors compared three paths for 2024 to 2035: business-as-usual, moderate transition (25% renewable share) and aggressive decarbonization (50% renewable share, full liquid immersion cooling). They report a net positive result only when large renewable-energy procurement, best-practice cooling and transparent verification of AI benefits all happen together.

The stakes are regional: GCC grids run at roughly 430 to 650 gCO2/kWh, summer temperatures exceed 45 °C, and all six countries sit in the highest baseline water-stress category. The Net Impact Score is a scenario-conditional decision aid rather than a causal forecast, because AI-benefit attribution and scaling depend on assumptions and limited operator data.

Main points

  1. The model covers three scenarios to 2035: business-as-usual (8% annual capacity growth), moderate transition (10%) and aggressive decarbonization (10%).
  2. The aggressive scenario assumes a 50% renewable share, a 45% cut in grid carbon intensity and a 30% PUE improvement; the business-as-usual case assumes a renewable share of about 5 to 8%.
  3. Weights in the scoring: carbon 0.40, water 0.25, social 0.20 and governance 0.15, derived from GCC climate pledges and checked with three experts.
  4. Sensitivity tests found renewable energy share and the AI benefit scaling factor had the greatest effect on the net score; social and governance dimensions mattered less.
  5. Cooling is a major load: 30 to 40% of facility energy without next-generation cooling, and GCC conventional operations run at a PUE of 1.42 against a best practice of 1.09.
  6. Indirect water use through desalination is estimated at 1.5 to 2.0 times the direct water used for cooling.
  7. Benefits counted only if they were in the GCC, independently verified, additional to business-as-usual and directly attributable to AI; six of twelve candidate case studies qualified.

The full story

The study uses secondary data from peer-reviewed papers, IEA, UNFCCC and Uptime Institute databases and independently verified case studies, because GCC operators treat energy, water and emissions data as confidential. It applies a linear weighted model with min-max normalization, and the Net Impact Score is the weighted benefit score minus the weighted footprint score.

Scenario inputs differ sharply. Business-as-usual keeps standard air cooling, minimal efficiency gains and no scaling of AI benefits. The moderate path adds hybrid air and liquid cooling and 50% of AI benefit potential, while the aggressive path assumes full liquid immersion cooling, 85% utilization and 100% of potential.

Background evidence shows why the region is hard. Saudi Arabia's grid is about 620 gCO2/kWh and Kuwait's about 650, against a global average near 440. Reverse osmosis desalination uses 3.5 to 4.5 kWh per cubic metre, so cooling water carries its own energy and carbon cost.

On the benefit side, the paper cites AI-enabled building HVAC optimization with 20 to 40% energy reductions, petrochemical pilots with 5 to 15% lower energy intensity, and smart grids that could cut solar and wind curtailment by 15 to 25%. It notes most such evidence comes from small pilots, and that Dubai's 8.5% per-capita electricity drop from 2019 to 2023 cannot be attributed clearly to AI.

What this doesn’t fix

The score is a scenario-conditional indicator based on secondary data and assumed weights, not an absolute causal prediction; AI-benefit attribution, scaling and the capacity figures depend on assumptions and limited operator data.

Sources

1 on file
  1. Frontiers in Environmental Sciencedoi:10.3389/fenvs.2026.1852462
    Peer-reviewed2026-07-15

Talk it over

Four questions to start onThis record found a gain and a problem in the same place. Which matters more to you, and who might answer differently?

Start with a question

Gulf AI data centers: model finds emissions rise by 2035 and AI offsets fall shorthttps://truvace.com/item/net-sustainability-assessment-of-ai-driven-data-centers-in-the-gcc-an-mcda-appro

  1. This record found a gain and a problem in the same place. Which matters more to you, and who might answer differently?

    The gain A scenario model of GCC data centers suggests AI applications in buildings, industry and utilities could offset part of the footprint, and net-positive outcomes appear only when renewable procurement, best-practice cooling and verified AI benefits combine.

    The problem A scenario model of GCC data centers suggests that under business-as-usual high growth, emissions rise substantially by 2035, with AI sustainability applications offering only partial offsets even under optimistic adoption assumptions.

    Answer in the thread

  2. How do we know? What can this kind of evidence show, and what can’t it?

    The evidence This rests on a benchmark or lab test, from Frontiers in Environmental Science.

    Answer in the thread

  3. What does it leave open? What else would you want to know before acting on it?

    The record’s own caveat The score is a scenario-conditional indicator based on secondary data and assumed weights, not an absolute causal prediction; AI-benefit attribution, scaling and the capacity figures depend on assumptions and limited operator data.

    Answer in the thread

  4. Have you seen AI used this way where you study, work or live? Did it go the same way?

    Answer in the thread

The questions are the same on every record. The details under them are taken from this record as written.

What readers say

More from Climate