Greenhouse gases from data center boom could outpace entire nations

2026-04-23 · Source: AI - Ars Technica · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure · Depth: Fundamental Awareness, long

Summary

New natural gas projects designed to power just 11 data center campuses across the US could generate over 129 million tons of greenhouse gases annually, exceeding Morocco's 2024 emissions. These projects, supporting major AI companies like OpenAI, Meta, Microsoft, and xAI, often utilize "behind-the-meter" power generation to bypass grid constraints and public resistance to higher energy bills. Examples include xAI's Colossus campuses in Tennessee and Mississippi, projected to emit over 6.4 million tons of CO2 equivalents each, and a Chevron-backed project for Microsoft in West Texas, potentially emitting 11.5 million tons annually. While permitted emissions represent a theoretical maximum, data center power demands are constant, suggesting actual emissions could be closer to these high projections. The Stargate Project, a multi-company AI effort, and other massive campuses like Fermi's and Pacifico Energy's, also show potential for tens of millions of tons of annual emissions, challenging corporate carbon reduction pledges.

Key takeaway

For CTOs and VPs of Engineering evaluating AI infrastructure expansion, recognize that the rapid build-out of data centers, particularly with behind-the-meter natural gas power, carries substantial and often underestimated greenhouse gas implications. Your teams should scrutinize energy sourcing strategies beyond grid connections, factoring in the long-term environmental and regulatory risks associated with direct fossil fuel generation, even if initially presented as a "clean" or "bridge" solution.

Key insights

The AI boom's data center expansion is driving a significant, under-recognized surge in natural gas-powered, behind-the-meter emissions.

Principles

• Behind-the-meter power bypasses grid constraints.
• Data center power demand is consistently high.
• Permitted emissions often exceed actual output.

In practice

• Track behind-the-meter gas permits.
• Evaluate data center energy portfolios.
• Consider nuclear energy for AI infrastructure.

Topics

• Data Center Emissions
• Behind-the-Meter Power
• AI Infrastructure
• Natural Gas Power Plants
• Corporate Sustainability

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Tech Journalist, Executive, Policy Maker

Related on AIssential

• Officials hugely underestimated impact of AI datacentres on UK carbon emissions - The Guardian — UK government vastly underestimated AI datacentre carbon emissions, raising climate concerns.
• The AI Boom Is Taking Over the U.S. Power Grid, and This Map Shows Where It’s Happening — The AI boom is rapidly expanding data center infrastructure, placing immense strain on the U.S. power grid and energy resources.
• Assessing the Carbon Emissions and Energy Consumption of U.S. Hyperscale Data Centers — US hyperscale data centers disproportionately rely on carbon-intensive electricity, driving significant emissions and energy demand.
• Data centers and crypto could hike power costs by 57% by 2030 — Data center and cryptocurrency expansion threatens to reverse U.S. power sector gains, driving up costs and CO2 emissions by 2030.
• What It Will Take to Make AI Sustainable — AI's rapid growth is increasing fossil fuel-powered data centers, challenging sustainability goals.
• Powering AI Growth: Google And TotalEnergies In Malaysia — Hyperscale operators are directly securing renewable energy to meet AI-driven data center power demands, circumventing grid bottlenecks.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI - Ars Technica.