UN warns AI boom carries severe and mounting environmental costs
Data centres already use as much energy as France, and the UN warns the situation will worsen significantly
The rapid global expansion of artificial intelligence is set to impose significant and largely underappreciated environmental costs on humanity in the years ahead, the United Nations has cautioned in a stark new report published on Wednesday.
The comprehensive study, produced by the United Nations University Institute for Water, Environment and Health, shines a light on the considerable strain that the AI industry's accelerating growth is placing on the planet's natural resources — strain that experts warn is routinely overlooked in mainstream discussions of the technology.
Energy consumption on a staggering scale
According to the report, data centres and AI models are already reliant on enormous quantities of resources, and that demand is set to intensify considerably.
By 2030, AI could account for nearly 3 per cent of the world's total electricity consumption. To contextualise the current scale, data centres are already consuming energy at a level comparable to the entire nation of France.
In 2025 alone, data centres equipped with servers and cooling systems used an estimated 448 terawatt-hours of electricity.
Electronic waste piling up
The report also flags the electronic waste crisis that AI infrastructure is generating. By 2030, the volume of e-waste produced by the AI sector could reach the equivalent of 250 Eiffel Towers annually — a figure that underscores the physical and material cost of the technology's expansion.
A mounting water crisis
The water demands of AI infrastructure are equally alarming. In 2025, data centres consumed approximately 9.3 trillion litres of water for cooling purposes alone. The report projects that by 2030, AI's water consumption could be sufficient to meet the drinking needs of all 8.1 billion people on Earth for more than one and a half years.
The renewable energy paradox
The report also highlights what it describes as a renewable paradox — the reality that switching to alternative energy sources such as bioenergy may reduce carbon emissions but can simultaneously increase water and land usage by a factor of up to 100.
"Low-carbon is not automatically low-water or low-land, and evaluating sustainability through a single metric can hide trade-offs and shift burdens onto places already facing water stress or land pressure," the report states.
The burden falls on the most vulnerable
Professor Te Taka Keegan of the AI Institute at the University of Waikato offered a sobering assessment of who ultimately bears the consequences of this environmental toll.
"The environmental burden falls hardest on communities least likely to capture the benefits. As AI is embedded into everyday platforms and switched on by default whether users choose it or not, that footprint compounds at scale," he said.
A call for sustainable AI development
UN researchers are urging governments and technology companies to factor water and energy requirements into the planning and construction of AI infrastructure and data centres from the outset. The central message of the report is that technological advancement must always remain "environmentally manageable and sustainable."