A new report from the United Nations University, released on 3 June, delivers a stark warning: the environmental footprint of global data centres already rivals that of entire nations, and the rapid expansion of artificial intelligence is set to double their energy, water, and pollution impact by 2030. The study, the first comprehensive global assessment of its kind, finds that data centres consumed 448 trillion watt-hours of electricity last year—more than all but ten countries—and produced roughly 189 million tonnes of carbon dioxide, equivalent to Argentina's annual emissions.
Kaveh Madani, a water scientist and director of the UN University Institute for Water, Environment and Health in Canada, and a co-author of the report, puts it bluntly: “If you look at these numbers, we're seeing scales comparable to nations. The demand is enormous.” By 2030, data centres could account for nearly three percent of the world's projected electricity use, with 935 trillion watt-hours. If they were a country, they would rank sixth in global power consumption, generating nearly 399 million tonnes of CO₂. The report focused on energy but noted that water used for cooling is also massive—about 4.5 trillion litres in 2024.
The surge is driven largely by AI, which currently accounts for about 20 percent of data centre energy use but is expected to reach 40 percent by 2030. A typical ChatGPT-style query consumes roughly 200 times more energy than a basic email spam filter, and generating images or video requires even more. Training models like GPT-3 used about 1.3 billion watt-hours; its successor, GPT-4, used 50 to 70 billion watt-hours. The report warns that without intervention, the environmental cost will escalate sharply.
How Users Can Cut Their AI Impact
Madani offers a practical, if provocative, suggestion: be less polite. “If you're too polite, then that extra ‘please’ you put there can make a huge difference,” he says. “You've got to be very precise and be short.” A 2024 survey by British publisher Future found that 70 percent of people are polite to AI, with 55 percent saying it's “the nice thing to do” and 12 percent joking about avoiding robot uprisings. But the report calculates that cutting word use by 30 percent can reduce AI energy consumption by 25 percent—saving roughly the same amount of electricity as 700,000 people in Africa use in a year.
The report's significance extends beyond its numbers. Fengqi You, a Cornell University energy engineering professor who directs AI sustainability research, notes that “its value is that a UN institution is putting carbon, water, land, life-cycle impacts and environmental justice into one frame” for an issue often shrouded in secrecy. Jean Su, director of the Energy Justice Program at the Center for Biological Diversity, calls it the first global report “that shines a light on the environmental harms of AI.”
Industry representatives defend the sector's progress. Caleb Max, president of the National Artificial Intelligence Association, argues that “AI is rapidly becoming part of our everyday lives and adding benefits that improve safety, help people live longer, work more efficiently, enhance food production, and reduce poverty. The evidence is growing daily that the energy return on investment of AI development is transformative.” Josh Levi, president of the Data Center Coalition, says the industry is “committed to working with policymakers, local communities, and industry partners to ensure that as data centres grow, they do so responsibly, transparently, and in ways that reflect the best available practices.”
For European readers, the implications are direct. Data centre expansion is accelerating across the continent, from energy corridors linking the Caspian to Europe to new facilities in the Netherlands, Ireland, and Germany. The EU has offered limited fiscal flexibility on energy costs in response to member states' demands, as seen in Brussels' response to Italy. Meanwhile, the EBRD has cut growth forecasts amid energy shocks, underscoring the fragility of Europe's energy landscape. The report arrives just after Monterey Park, California, became the first US city to permanently ban new data centres—a move that could inspire similar debates in European municipalities.
Madani stresses that AI's impact is not virtual. “A lot of hardware is behind all these operations that to us seem very, very clean because we don't see smoke out of our devices. On our cellphone, there is no visible smoke. But somewhere else someone is suffering.” The message is clear: as AI becomes embedded in daily life, users and policymakers alike must confront its physical costs—and perhaps start by dropping the ‘please.’
