OpenAI's Altman Dismisses Data Center Water Use Concerns
The CEO claims AI facilities consume no more water than office buildings, but experts say the data to verify such claims doesn't exist.

OpenAI CEO Sam Altman is pushing back against growing public concern over AI data centers' water consumption, characterizing the issue as a "meme" unsupported by facts. Speaking on the Sources podcast with Alex Heath, Altman argued that modern large data centers consume no more water than typical office buildings.
Altman offered a specific comparison to illustrate his point: it takes approximately 38,000 ChatGPT queries to consume as much water as producing a single California almond. He has previously stated that a typical ChatGPT query requires roughly 0.32 milliliters of water, while a 2019 study by U.S. Geological Survey-affiliated researchers found that growing one California almond requires an average of 3.56 liters of water.
Why it matters
The dispute over data center water consumption arrives as AI infrastructure expands rapidly and communities increasingly resist new facilities. Without transparent, standardized reporting requirements, the industry faces a credibility gap at precisely the moment it needs public acceptance for major infrastructure investments. California legislation now pending could establish the disclosure framework needed to move the debate from speculation to verifiable facts.
Experts say data is insufficient
Researchers dispute the possibility of making definitive comparisons with current information. Shaolei Ren, a professor of electrical and computer engineering at UC Riverside, explained that water consumption at any facility depends on numerous interacting variables including geographic location, ambient temperature, cooling technology, and query complexity. "The information we have is so limited," Ren said.
Michael Kiparsky, who leads the Wheeler Water Institute at UC Berkeley's Center for Law, Energy, and the Environment, pointed to the lack of published data. "There's no published data — certainly not for California — and what is public seems all over the place," he noted.
Disclosure legislation advances
Two bills requiring data centers to disclose their water use have cleared the California legislature this month and await Governor Gavin Newsom's decision. Newsom vetoed similar legislation last year. The pending bills signal recognition that baseline transparency is currently absent from the sector.
Public opposition intensifies
Communities nationwide are increasingly resisting new data center projects. A Gallup survey released in May found that 71% of Americans opposed having a data center built in their community, with 70% citing environmental impact as their reason.
Altman acknowledged the industry must address public skepticism. "I think the right way to get people to like something is to deliver them value," he said on the podcast.
The tension between AI companies' infrastructure needs and community concerns over resource consumption appears likely to intensify as the industry scales. Without standardized disclosure requirements, claims and counterclaims about water use will remain difficult to verify independently.
These details were first reported by Business Insider and CalMatters.
This is an original analysis by the Omega editorial team. Source reporting: AI Watch.
Want systems like this working for your business?
Book a Call

