AI Growth Drives Data Centre Emissions Surge

The rapid expansion of artificial intelligence is accelerating electricity demand and carbon emissions among the world’s biggest tech giants, with a new report from the United Nations’ International Telecommunication Union (ITU) sounding the alarm over the environmental consequences of unchecked digital growth. From 2020 to 2023, indirect carbon emissions from AI-linked data centre operations at companies like Amazon, Microsoft, Alphabet (Google’s parent company), and Meta surged by an average of 150%, driven primarily by their growing reliance on energy-intensive infrastructure.

Amazon led the pack with a dramatic 182% spike in indirect emissions during the three-year period, followed by Microsoft at 155%, Meta at 145%, and Alphabet at 138%. These emissions fall under the scope of indirect or Scope 2 emissions, which include greenhouse gases generated from purchased electricity, heating, cooling, and steam—services that are essential to keeping massive data centres operational around the clock. These facilities, which power everything from cloud storage to AI model training, are increasingly energy-hungry and carbon-intensive, making them a critical point of concern for climate advocates and policymakers alike.

The ITU report draws attention to the scale and speed of this shift, noting that electricity use by data centres is growing four times faster than the overall global increase in electricity consumption. As the demand for AI applications like generative language models, real-time translation tools, and advanced image processing continues to escalate, the infrastructure that supports them is struggling to keep pace in a sustainable way.

This explosive growth has raised serious concerns about the climate impact of the digital economy. The ITU estimates that, without aggressive mitigation measures, AI-linked emissions could soar to 102.6 million metric tons of carbon dioxide equivalent (CO₂e) annually. Such levels would make the sector one of the world’s major industrial emitters, comparable to the footprint of some mid-sized countries. For nations and companies aiming to meet ambitious net-zero goals, this trajectory poses a significant threat.

The report reviewed the emissions performance of 200 major digital firms and found that despite public commitments to sustainability, the actual emissions outcomes remain largely unchanged. Many tech companies have made bold pledges to achieve net-zero emissions within the next two decades, but the ITU notes a growing gap between stated targets and real-world progress.

Some of the firms under scrutiny have responded by pointing to their ongoing sustainability efforts. Meta, for instance, highlighted its sustainability report, in which it outlines ongoing work to cut emissions, energy use, and water consumption across its global data centre operations. Amazon emphasized its investment in cleaner energy sources, including nuclear and renewables, as part of its broader commitment to powering operations sustainably. Microsoft also stressed its efforts to curb energy use, citing a doubling in power savings over the past year and innovations such as chip-level liquid cooling systems designed to reduce thermal waste in its data centres.

However, the ITU report suggests that such initiatives, while commendable, remain insufficient to offset the broader emissions surge tied to AI growth. It warns that unless digital leaders implement more transformative changes—such as energy efficiency mandates, green cloud computing standards, and stricter emissions disclosures—the sector risks derailing long-term environmental and business objectives.

The disconnect between technological advancement and environmental responsibility has created a growing sense of urgency among policymakers, investors, and sustainability advocates. C-suite leaders are now being challenged to rethink their infrastructure strategies, not only to meet stakeholder expectations but also to secure regulatory approval and maintain social license amid increasing public scrutiny.

The challenge is clear: AI and digital innovation must be decoupled from carbon-intensive growth. This will require not only improved efficiency but also a systemic shift toward green infrastructure, smarter computing architecture, and greater transparency in carbon accounting.

As digital transformation continues to shape the global economy, the environmental cost of powering that transformation cannot be ignored. The ITU’s findings serve as a wake-up call for the tech industry to align its innovations with planetary boundaries—or risk facing the consequences of unsustainable growth.