OpenAI’s Altman Calls Data Center Water Concerns “Fake” as AI Infrastructure Costs Draw Scrutiny

OpenAI's Altman Calls Data Center Water Concerns "Fake" as AI Infrastructure Costs Draw Scrutiny

OpenAI CEO Sam Altman dismissed criticisms of artificial intelligence's environmental footprint on Friday, calling concerns about data centers' water consumption "fake" and defending AI's energy demands by comparing them to human resource use.

Speaking on the sidelines of the India AI Impact summit in New Delhi, Altman told The Indian Express that claims circulating online about ChatGPT using gallons of water per query were "completely untrue, totally insane" with "no connection to reality." The remarks come as finance leaders increasingly scrutinize the infrastructure costs and sustainability implications of AI deployments across their organizations.

The comments represent a rare public defense of AI's resource intensity from a major industry leader, even as data center operators face mounting pressure from investors and regulators over environmental impacts. Data centers traditionally rely on large volumes of water to cool electrical components and prevent overheating—a process that has drawn criticism as AI workloads expand.

Altman acknowledged that total AI energy consumption is rising and urged the industry to pursue cleaner power sources. However, he argued that the comparison itself was misguided, noting that humans also consume energy to carry out tasks—an implicit suggestion that AI's resource use should be evaluated against the work it displaces rather than in isolation.

The debate over AI infrastructure costs has moved from academic circles to boardrooms as CFOs weigh the financial and reputational risks of expanding AI capabilities. While some newer data centers have adopted cooling technologies that eliminate water use entirely, many active facilities still depend on it. A January 2026 report from water technology company Xylem and Global Water Intelligence projected that water drawn for data center cooling would more than triple over the next 25 years as computing demands grow.

For finance leaders, Altman's framing presents a challenge: how to evaluate AI investments when the full cost—including environmental externalities—remains contested. The OpenAI chief's dismissal of water concerns as fabricated may complicate efforts by sustainability-focused investors and ESG teams to quantify the true resource footprint of AI initiatives.

The timing of Altman's comments is notable. As AI moves from pilot projects to production deployments, finance teams are discovering that infrastructure costs can dwarf initial software licensing fees. The question of whether those costs include meaningful environmental liabilities—and whether executives like Altman are accurately characterizing them—will likely influence capital allocation decisions in the months ahead.

What remains unclear is whether Altman's characterization of water consumption claims as "insane" reflects genuinely misleading viral statistics or a broader pushback against scrutiny of AI's resource intensity. For CFOs evaluating AI vendors, the answer matters: it's the difference between a messaging problem and a material cost that someone, eventually, will have to pay.