Data center emissions 57% higher than expected as AI boom drives unprecedented carbon footprint

Data Center Emissions Far Exceed Previous Estimates

Data centers emitted 286 million tonnes of CO₂ in 2025, according to a new study by Allianz Trade, marking a stark 57% increase over estimates from the International Energy Agency1

1

2

2

. The sprawling, power-hungry facilities used to store critical IT infrastructure like servers are being built worldwide as the AI boom drives unprecedented demand for computing power. This revelation underscores a growing environmental crisis tied directly to artificial intelligence expansion, with nearly 70% of global data center emissions currently concentrated in the U.S. and China, the world's AI leaders2

2

.

AI Workloads Driving Explosive Growth in Data Center Electricity Consumption

AI already accounts for between 15% and 20% of data center electricity consumption, and this share could climb to 40% by 2030, the report warns1

1

. "Data centres are evolving from a marginal factor into a structural driver of electricity demand in many regions," said Patrick Hoffmann, senior climate economist at Allianz2

2

. The AI-driven environmental impacts extend beyond carbon emissions, creating cascading effects across multiple resource systems. Without immediate decarbonization efforts to transform power grids, data center emissions would more than double by 2030, leading to an estimated $154 billion in annual climate damages—up from $68 billion today1

1

.

Geographic Disparities Highlight Importance of Cleaner Energy Sources

The same computing power can generate vastly different emission levels depending on electricity sources. In India, emissions linked to electricity exceed 600 grams of CO₂ per kilowatt-hour, compared with less than 30 grams in Norway or Sweden, where power generation relies heavily on decarbonizing power grids2

2

. This disparity highlights how location and energy infrastructure choices can dramatically alter the carbon footprint of identical AI workloads. Climate damage related to AI workloads alone could exceed $50 billion by 2030, emphasizing the urgent need for tech companies to prioritize cleaner energy sources in their expansion strategies1

1

.

Strain on Water Resources Adds to Environmental Concerns

Beyond carbon emissions, data centers are creating immense strain on water resources. The facilities could require 1.3 trillion to 1.8 trillion liters of water by 2030—comparable to Switzerland's annual water consumption1

1

. This water consumption primarily supports cooling systems essential for maintaining optimal operating temperatures in these power-intensive facilities. As AI applications continue gobbling ever-greater computing power, the dual challenge of carbon emissions and water consumption demands coordinated action from policymakers, technology companies, and energy providers to mitigate long-term environmental damage.