The United Nations has issued a stark warning about the environmental impact of artificial intelligence, projecting that by 2030, AI could consume nearly 3 per cent of the world's electricity, produce carbon emissions comparable to the United Kingdom's total 2025 output, use enough water to meet the drinking needs of every person on Earth for more than a year and a half, and generate electronic waste equivalent to discarding 250 Eiffel Towers each year.
Comprehensive Assessment of AI's Environmental Costs
These findings come from a new report published on Wednesday by the United Nations University Institute for Water, Environment and Health, which researchers describe as the most comprehensive assessment yet of AI's environmental footprint. While previous discussions have focused heavily on carbon emissions, the report argues that this is only part of the story. Cutting emissions alone may not significantly reduce AI's overall environmental harm.
“Low-carbon is not automatically low-water or low-land,” the report states, “and evaluating sustainability through a single metric can hide trade-offs and shift burdens onto places already facing water stress or land pressure.”
Data Centre Energy Consumption Soaring
Data centres, the vast warehouse-like facilities filled with servers and cooling systems that run continuously to power AI, consumed an estimated 448 terawatt-hours of electricity in 2025, roughly equivalent to the entire national consumption of France. A terawatt-hour is one billion kilowatt-hours, the unit used on household electricity bills. AI workloads accounted for around 20 per cent of that total. If that share rises to the expected 40 per cent by 2030, AI-related electricity use could reach 374 terawatt-hours. On current trajectories, the report projects that total data centre energy use could roughly double to 945 terawatt-hours, enough to power all 1.3 billion people in sub-Saharan Africa for over five years. The land required to generate that electricity would exceed 14,000 square kilometres, roughly the area of Northern Ireland.
Water Consumption Strains Resources
The water consumed in cooling this infrastructure adds another challenge. Data centres used an estimated 9.3 trillion litres in 2025, a figure that the report says would meet the drinking water needs of the planet's 8.1 billion people for more than a year and a half. Even where some of that water is returned to the environment, large-scale withdrawals strain aquifers and river systems, particularly in regions already facing water scarcity. In the Netherlands, a large data centre drawing heavily on water supplies during a drought year prompted opposition from local farmers.
Training vs. Daily Use: Environmental Costs Shift
Training a single large AI model such as ChatGPT-5 requires around 100 gigawatt-hours of electricity, equal to the annual residential power consumption of 770,000 people in sub-Saharan Africa, along with an estimated one billion litres of water and a land footprint covering roughly 215 football fields. However, the report found that the environmental cost of training, though large, has been overtaken by the cost of daily use. ChatGPT alone processes an estimated 2.5 billion prompts per day. A conventional Google search uses about 0.3 watt-hours of electricity, while an AI-enhanced generative search uses up to 3 watt-hours, a tenfold increase, applied across an estimated 5 trillion searches a year.
User Choices Matter
The choices users make affect these numbers more than is widely understood. Switching to a concise response mode can reduce ChatGPT's output by 30 per cent, saving 87 to 98 gigawatt-hours of electricity per year, equivalent to the annual residential electricity of nearly 760,000 people in sub-Saharan Africa. Removing pleasantries, such as not saying please or thank you, makes prompts more concise and reduces the cumulative footprint at scale.
AI Video Generation a Growing Concern
The rising popularity of AI-generated videos is becoming a major environmental concern, the study says. A single high-resolution AI video clip requires more than 415 watt-hours of electricity, more than generating hundreds of AI images. AI videos have also been improving rapidly in quality. But as resolution and frame count increase, energy requirements also rise exponentially. Video generation has become embedded in mainstream social media platforms, with sites encouraging users to create and post more AI videos as part of viral trends. The report warns that this is becoming an infrastructure-scale problem.
Professor Alistair Knott of the Centre for Data Science and AI at Victoria University of Wellington, who was not involved in the report, said that while the study calls out growing investments by AI companies, it fails to note that AI companies depend on increased growth of the AI market for their own survival. “The only way companies can survive is to grow the market for AI products at an ever-increasing pace, but that’s not necessarily what the world needs,” he said. “Governments, elected by citizens, are better placed to make the right decisions about how much AI we need, and to trade this need off against environmental impacts.”
Renewable Energy Not a Panacea
The report found that powering data centres with renewable energy does not automatically make them sustainable. Switching from coal to bioenergy can reduce the carbon footprint of electricity generation by 72 per cent, but the water footprint of bio-energy is on average more than 30 times that of coal, and its land footprint is 100 times as great. For example, Brazil's hydro-powered grid produces electricity 77 per cent below the global carbon average, but its water and land footprints are nearly triple the global mean.
Infrastructure Growth Outpacing Planning
In Ireland, data centres now account for 21 per cent of the country’s total metered electricity, up from 5 per cent a decade ago, exceeding all urban household consumption combined. Researchers say this results from AI infrastructure growth outpacing energy planning. The national grid operator has paused new approvals around Dublin until 2028. Professor Te Taka Keegan of the AI Institute at the University of Waikato, also not involved in the report, said the concentration of infrastructure raised environmental justice concerns. “The environmental burden falls hardest on communities least likely to capture the benefits,” he said. “As AI is embedded into everyday platforms and switched on by default whether users choose it or not, that footprint compounds at scale.”
Call for Government Action
The UN researchers urge governments to start factoring AI infrastructure into water and energy planning. They also recommend that tech companies include environmental considerations when planning new features. “Technological advancement must remain environmentally manageable,” the report’s authors write. “Real progress depends on embedding sustainability at every level, from hardware and model design to deployment, governance, and public use.”
