Sam Altman would like to remind you that humans use a lot of energy, too

OpenAI CEO Sam Altman addressed questions about AI’s environmental footprint this week during an appearance at an event hosted by The Indian Express, pushing back on several popular claims while acknowledging broader concerns about overall power demand.

Altman — who was in India for a major AI summit — said that much of the online discussion around AI’s water usage is misleading. He called the idea that ChatGPT consumes huge amounts of water per prompt “totally fake,” while noting that water use was once a legitimate issue when data centres relied on evaporative cooling.

“Now that we don’t do that, you see these things on the internet where, ‘Don’t use ChatGPT, it’s 17 gallons of water for each query’ or whatever,” Altman said. “This is completely untrue, totally insane, no connection to reality.”

At the same time, Altman said it’s reasonable to be concerned about AI’s energy consumption — not necessarily on a per-question basis, but because usage at scale is growing rapidly across the world. In his view, the situation underscores a larger need for the energy sector to shift faster toward low-carbon sources.

He said it’s “fair” to worry about “the energy consumption — not per query, but in total, because the world is now using so much AI,” adding that the world needs to “move towards nuclear or wind and solar very quickly.”

Altman’s comments come as researchers attempt to estimate AI’s resource usage amid a lack of transparency requirements. There is currently no legal obligation for major technology companies to publicly disclose detailed figures on the water and electricity consumption associated with their AI systems, so scientists have tried to analyse the impact independently. At the same time, large-scale data centre expansion has been linked in some places to rising electricity costs.

During the discussion, the interviewer referenced a previous conversation with Microsoft co-founder Bill Gates and asked whether it’s accurate to claim that a single ChatGPT query uses about the same energy as 1.5 iPhone battery charges. Altman rejected that characterisation outright.

“There’s no way it’s anything close to that much,” he said.

Altman also argued that public debates about AI energy use are often framed in a way he considers unfair — particularly when they focus heavily on the energy required to train large models, without comparing that to the energy costs involved in producing a human capable of answering similar questions.

He criticised comparisons that measure “how much energy it takes to train an AI model, relative to how much it costs a human to do one inference query.” Then he made his own counterpoint: training a person requires years of energy intake and a long chain of biological and societal development.

“But it also takes a lot of energy to train a human,” Altman said. “It takes like 20 years of life and all of the food you eat during that time before you get smart. And not only that, it took the very widespread evolution of the 100 billion people that have ever lived and learned not to get eaten by predators and learned how to figure out science and whatever, to produce you.”

From Altman’s perspective, the more meaningful comparison is what happens after an AI model is already trained: how much energy it takes for a system like ChatGPT to answer a question versus the energy required for a human to do the same work.

So, he said, the fair framing is: “If you ask ChatGPT a question, how much energy does it take once its model is trained to answer that question versus a human? And probably, AI has already caught up on an energy efficiency basis, measured that way.”

A full recording of the interview is available to watch, and the segment on water and energy usage begins at approximately 6:35.