The United Nations’ International Telecommunication Union (ITU) reports that carbon emissions from major tech companies rose by 150% on average between 2020 and 2023, driven by advancements in AI and increased energy use in data centers.

Amazon’s emissions grew by 182%, Microsoft’s by 155%, Meta’s by 145%, and Alphabet’s by 138%, covering both direct operations and energy consumption, according to the ITU.

The ITU highlighted that the growing use of AI and cloud computing has significantly raised electricity consumption.

Data centers, essential for digital services, have experienced a 12% annual increase in electricity use since 2017, according to the International Energy Agency (IEA).

In 2024, data centers consumed 415 terawatt-hours (TWh) of electricity, accounting for 1.5% of global power demand.

If the trend continues, data centers could consume 945 TWh by 2030, surpassing Japan’s total power use, while just 10 tech companies make up over half of this energy demand.

🔗 Source: Al Jazeera

🧠 Food for thought

1️⃣ The tech emissions surge fits into a longer growth pattern

The 150% rise in tech companies’ operational carbon emissions isn’t an isolated AI phenomenon but part of a longer trend in the digital sector’s growing environmental impact.

The tech industry was already projected to contribute 3.0-3.6% of global greenhouse emissions by 2020, comparable to the aviation industry, well before the current AI boom began 1.

This context matters because the 297 million tonnes of CO₂ equivalent emitted by these companies in 2023 equals the combined emissions of Argentina, Bolivia, and Chile 2, showing the substantial scale of tech’s climate impact.

The surge continues despite efficiency improvements in data center technology, illustrating what experts call the “rebound effect,” where efficiency gains are outpaced by rapid expansion of digital services and computing power 1.

Understanding this pattern helps explain why technological improvements alone haven’t reduced the sector’s overall environmental footprint, as total energy demand continues to outstrip efficiency gains.

2️⃣ AI emissions face a significant regulatory vacuum

Unlike other high-emission sectors, AI’s rapidly growing carbon footprint operates in a regulatory environment with minimal oversight or standardized reporting requirements.