The United Nations University Institute for Water, Environment and Health published a report on June 3, 2026, warning that artificial intelligence’s expanding infrastructure could consume as much electricity by 2030 as nearly triple the combined annual usage of Pakistan, Bangladesh and Nigeria. The report, titled “Environmental Cost of AI’s Energy Use: Carbon, Water and Land Footprints,” was produced by UNU-INWEH researchers based in Richmond Hill, Ontario, Canada. It calls on governments and technology companies to disclose the full environmental costs of artificial intelligence rather than measuring impact through carbon emissions alone.
Global data centers powering AI could consume 945 terawatt-hours of electricity annually by 2030, according to the report. That figure compares with 448 terawatt-hours consumed by data centers worldwide in 2025, a level that would have ranked as the eleventh-largest national electricity consumption in the world, behind France and ahead of Saudi Arabia, the report states.
The water demands tied to that electricity use carry separate consequences. UNU-INWEH researchers project that AI-related water consumption could equal the basic annual domestic water needs of 1.3 billion people in Sub-Saharan Africa by the end of the decade. The land footprint associated with power generation and supply chains for AI infrastructure may exceed 14,500 square kilometers, according to the report.
Dr. Miriam Aczel, a UNU-INWEH researcher and the report’s lead author, said the findings challenge assumptions that renewable energy automatically makes AI infrastructure environmentally sound. “What surprised us most is how often the choices that look greenest from a carbon perspective end up worse for water or for land,” she said. “If we keep judging AI sustainability by carbon alone, we might think that renewables make AI infrastructure clean but that is solving one problem while creating other problems, often in places that didn’t ask for it.”
The report’s authors stress that carbon, water and land footprints do not move in the same direction. A renewable energy source that lowers greenhouse gas emissions can simultaneously increase water consumption or land use, particularly in regions already facing resource scarcity, according to the findings.
Professor Kaveh Madani, director of UNU-INWEH and a lead investigator on the report, said the research is not intended to discourage AI development. “This report is not a case against artificial intelligence, a technological transformation that is improving the lives of billions of people around the world,” he said. “It is a call for using it responsibly and addressing its unintended impacts proactively.”
Madani, who was recently named the 2026 Stockholm Water Prize Laureate, also cautioned against relying on efficiency improvements to offset rising AI demand. Citing what economists call the rebound effect, the report argues that cheaper and more efficient AI tools tend to drive higher overall usage, increasing total resource consumption rather than reducing it.
The report identifies Ireland as a documented example of the strain AI infrastructure can place on national systems. Data centers accounted for 21% of Ireland’s total metered electricity consumption in 2023, exceeding usage by all urban households combined. Ireland’s national grid operator has paused new data center approvals around Dublin until 2028, according to the report.
Public attention has focused largely on the energy required to train large AI models, but the report finds that day-to-day usage of AI tools accounts for roughly 80 to 90 percent of total energy demand. One widely used AI service processes an estimated 2.5 billion prompts per day, consuming hundreds of gigawatt-hours of electricity annually, the report states. It also cites an estimate that an AI-enhanced internet search may consume roughly ten times more energy than a conventional search.
Beyond electricity and water, the report warns that AI infrastructure could generate up to 2.5 million tonnes of electronic waste annually by 2030.
The benefits and costs of AI’s global expansion are unevenly distributed, according to the report. Professor Tshilidzi Marwala, rector of the United Nations University and under-secretary-general of the United Nations, said this imbalance amounts to a governance failure rather than a technical limitation. “The global system building artificial intelligence must also govern it sustainably and fairly,” he said. “The concentrated development of AI infrastructure in the privileged areas of the world is creating a large digital divide that poses profound challenges in the equitable development of AI.”
For regions hosting AI infrastructure without sharing proportionally in its economic benefits, the report frames the issue as one of environmental justice. Countries that supply critical minerals and absorb electronic waste often do not capture the strategic or financial gains generated by AI services deployed elsewhere, according to the findings.
The report outlines six principles for what it calls a “responsible AI ecosystem”: transparency, efficiency by design, equity and environmental justice, lifecycle responsibility, global cooperation, and sustainable use. It recommends that governments integrate AI infrastructure planning into energy, water, and land-use policy, and that they require standardized environmental footprint reporting from AI developers.
Industry and AI developers are urged to treat model selection and default outputs as factors that determine environmental footprint, while data center operators and utilities are encouraged to apply cumulative impact assessments when deciding where to build. Investors are advised to treat electricity, carbon, water, and land footprints as material financial risks within AI infrastructure portfolios.
The global AI market is projected to grow from $189 billion in 2023 to nearly $4.8 trillion by 2033, according to figures cited alongside the report’s release.
No binding international mechanism currently requires AI companies to disclose environmental data tied to their models or data centers. The report calls for harmonized measurement standards across countries and for international institutions to help build computing capacity in regions currently excluded from AI infrastructure investment. UNU-INWEH researchers say further monitoring of electricity, water, and land use tied to AI expansion will continue as the 2030 projections approach.
