How Much Water Did AI Use in 2025? Study Compares It to Global Bottled Water Demand
Researchers estimated AI systems could consume hundreds of billions of liters of water, raising new questions about Big Tech’s hidden environmental costs.
A new study estimated that AI systems in 2025 could consume 312.5 to 764.6 billion liters of water, roughly comparable to the global bottled water industry’s annual demand. The research, published in Patterns by Alex de Vries-Gao of VU Amsterdam, also projected AI-related carbon emissions on par with New York City’s annual emissions.
The question is not just how much water AI uses. The question is who benefits from that resource use, who pays the cost, and whether Big Tech is being allowed to build a private empire on public infrastructure.
The Coffman Chronicle tracks how Big Tech, billionaires, and political power reshape everyday life. Subscribe free to follow this story as it develops
The findings raise conflict between the digital economy’s rapid growth and the physical limits of natural resources. As generative AI use has exploded this year, so too has demand for data centers that require water for cooling and electricity to run.
Researchers estimate AI’s water footprint in 2025 at 312.5–764.6 billion liters, roughly matching the global annual bottled water industry’s consumption when all sources are counted.
At the same time, AI-related carbon emissions are projected at 32.6–79.7 million metric tons of CO₂, a figure comparable to the total emissions attributed to New York City.
Subscribe free for daily political analysis they won’t broadcast. Join 110K+ readers →
Despite these startling numbers, there is significant uncertainty because major tech firms don’t publicly disclose AI-specific environmental data for electricity or water use.
“Significant uncertainty surrounds these figures,” the study notes, underscoring the need for better reporting.
These resource demands matter because they highlight the hidden environmental costs of AI’s growth, potentially shaping debates on sustainability, regulation, and energy planning.
Expect continued reporting and calls for corporate transparency and policy action as AI infrastructure expands.
What's next?
Water and carbon footprints may now be part of the AI conversation — not just computational horsepower.
Subscribe free for daily political analysis they won’t broadcast. Join 110K+ readers →