OpenAI CEO Sam Altman addressed concerns about AI’s impact on the environment at an event hosted by The Indian Express this week.
First off, Altman, who is in India for a major AI summit, said concerns about water use in AI were “completely bogus,” but acknowledged it was a real problem “back in the day when we were using evaporative cooling in data centers.”
“Now that we don’t do that, you see things on the Internet like, ‘Don’t use ChatGPT, it takes 17 gallons of water for each query,’” Altman said. “This is completely untrue, completely crazy and has nothing to do with reality.”
He added that because the world is using AI so heavily now, it’s “fair” to be concerned about “energy consumption in the aggregate, not just per query.” In his view, this means the world needs to “move very quickly towards nuclear or wind or solar.”
Because there is no legal requirement for technology companies to disclose how much energy and water they use, scientists have been trying to study this independently. Data centers are also linked to rising electricity prices.
Citing a previous conversation with Bill Gates, the interviewer asked if it was accurate to say that a single ChatGPT query is currently equivalent to 1.5 iPhone battery charges, to which Altman responded, “There’s nothing even close to that.”
Altman also complained that much of the discussion of ChatGPT’s energy use was “unfair.” This is especially true when focusing on “the amount of energy required to train an AI model compared to the cost of a human performing one inference query.”
Tech Crunch Event
Boston, Massachusetts
|
June 9, 2026
“But it also takes a lot of energy to train humans,” Altman said. “It takes about 20 years of life and all the food you eat during that period for you to become smart. And not only that, it took a very extensive evolution of 100 billion people learning how to avoid getting eaten by predators and figuring out science and things like that.”
So, in his view, a fair comparison would be: “If you ask ChatGPT a question, how much energy does that model need after it’s been trained to answer that question compared to a human? And perhaps AI has already caught up to energy efficiency standards and measured it that way.”
You can watch the full interview below. The conversation about water and energy use begins around 26:35.
