The United Nations warns that rapid artificial intelligence expansion could consume enough water to sustain the global population for over 1.5 years [1].
This projection underscores a growing tension between the acceleration of digital innovation and the preservation of critical natural resources. As AI models require immense computing power, the physical infrastructure supporting them places unprecedented pressure on global energy and water grids.
Released in May 2024, the UN report projects that AI could account for approximately three percent of world electricity use by 2030 [1]. This energy demand would generate carbon emissions comparable to those of the United Kingdom [1]. The report also highlights the production of unprecedented levels of electronic waste as hardware is rapidly cycled to keep pace with technological leaps [1].
These environmental costs stem from data centers that operate 24/7 to train and run complex models [2]. These facilities require massive amounts of electricity and water for cooling to prevent system failures. Michael Brown, writing for Techopedia, said data-center cooling can require up to 200 million gallons of water per day for AI workloads [4].
UN Climate Envoy Inger Andersen said if the world does not act now, AI could become one of the biggest water consumers on the planet [1]. The findings, released from UN headquarters in New York, suggest that the current trajectory of AI growth is unsustainable without significant intervention in how these systems are powered and cooled [1].
The report suggests that the intersection of energy and water needs creates a double-burden for regions already facing climate instability. While AI offers potential solutions for climate modeling, the immediate physical footprint of the technology may exacerbate the very crises it is intended to help solve [2].
“AI could become one of the biggest water consumers on the planet.”
The UN's findings signal a shift in the AI conversation from software capabilities to physical constraints. By quantifying the environmental cost in terms of national carbon footprints and global water supplies, the report suggests that the 'cloud' has a massive, tangible impact on earth. This likely necessitates new regulatory frameworks for data center efficiency and a move toward sustainable computing to avoid systemic resource depletion by 2030.
