AI Data Centers Could Match Water Use of 1.3 Billion People by 2030

Artificial intelligence infrastructure is on track to consume as much water as 1.3 billion people by the end of the decade, according to new research that reveals the technology's environmental impact extends far beyond carbon emissions.

A report from the United Nations University Institute for Water, Environment and Health (UNU-INWEH) quantifies the full scope of AI's resource demands, measuring not just carbon output but also water consumption for cooling and land use for energy infrastructure. The findings, first reported by TIME, paint a picture of competing environmental pressures that defy simple solutions.

The scale of AI's resource demands

By 2030, global data centers powering AI are projected to consume 945 terawatt-hours of electricity—nearly triple the combined annual electricity use of Pakistan, Bangladesh, and Nigeria, countries with a combined population exceeding 650 million people.

The water footprint is equally staggering. Data centers are expected to require water equivalent to the basic domestic needs of all 1.3 billion people in Sub-Saharan Africa for an entire year. Large facilities can consume up to 5 million gallons of water daily for server cooling.

The land footprint could exceed 5,590 square miles by 2030, roughly twice the size of the Jakarta metropolitan area, which currently houses more than 32 million people.

The renewable energy paradox

The report highlights a troubling trade-off: switching to cleaner energy sources can reduce carbon emissions while dramatically increasing other environmental impacts. Replacing coal with bioenergy cuts electricity's carbon footprint by 70 percent but increases water use more than 30-fold and land use 100-fold.

"What surprised us most is how often the choices that look greenest from a carbon perspective end up worse for water or for land," said Miriam Aczel, UNU-INWEH researcher and lead author of the report. "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."

Communities already feeling the impact

In 2025 alone, data centers consumed an estimated 448 terawatt-hours of electricity—more than Saudi Arabia's total consumption. In Ireland, data centers accounted for 21 percent of total metered electricity in 2023, exceeding urban household use. The country's national grid operator has paused new approvals around Dublin until 2028.

Water-stressed regions face particularly acute challenges. In Querétaro, Mexico, fast-tracked data center plans threaten water supplies amid prolonged droughts. Uruguay saw protests in 2023 when plans for a water-intensive data center emerged during a drought that made tap water unsafe to drink in the country's largest city.

Why it matters

The concentration of AI infrastructure creates a widening digital divide with environmental consequences. As of 2025, only 32 countries host AI-specialized data centers, with 90 percent of capacity concentrated in the United States and China. Meanwhile, AI infrastructure could generate up to 2.5 million metric tons of electronic waste annually by 2030, predominantly affecting low-income countries where waste is exported.

The report calls for permitting and environmental assessments that account for water and land use alongside carbon emissions. "We have a narrow window to ensure that the backbone of the technological revolution of our era develops within planetary limits," said Kaveh Madani, director of UNU-INWEH.

The findings were first reported by TIME and underscore the need for comprehensive environmental governance as AI infrastructure expands globally.