Sam Altman Dismisses ChatGPT Water Usage Allegations

OpenAI CEO Sam Altman pushed back against claims that AI systems consume excessive amounts of water, calling such assertions “fake” and disconnected from reality, according to CNBC. He made the remarks during an appearance at the India AI Impact Summit, where OpenAI also secured a 100MW AI data center deal with Tata.

Sam Altman Pushes Back on AI Water Consumption Criticism

Altman specifically rejected viral claims suggesting that ChatGPT uses large volumes of water for each query. He argued that the public debate often misunderstands how AI systems operate and exaggerates the environmental impact of individual interactions.

While dismissing concerns about water consumption per query, Altman acknowledged that AI’s overall energy demand represents a legitimate and growing issue. He said that as global AI adoption accelerates, electricity needs will continue to rise sharply.

Altman stressed the importance of expanding energy production to meet that demand. He pointed to nuclear, wind, and solar power as critical sources for supporting AI-driven infrastructure.

Comparing AI to Human Energy Use

During the discussion, Altman compared AI systems to human energy consumption. He noted that training a human brain over decades requires significant resources, suggesting that AI energy use should be evaluated within a broader context.

He also argued that critics often focus too heavily on the energy required to train large AI models. In reality, he said, inference, the process of generating responses from a trained model, consumes far less energy than initial training.

That comparison sparked debate online, with some technology leaders pushing back. Zoho co-founder Sridhar Vembu criticized the idea of equating human biological processes with technological systems.

Growing Scrutiny on Data Centers

Altman’s remarks come at a time when governments and environmental groups are closely examining the rapid expansion of AI data centers. Traditional facilities rely heavily on water for cooling, though newer designs increasingly use alternative cooling technologies.

Industry forecasts suggest that water demand for cooling could more than triple over the next 25 years as computing needs grow. Meanwhile, the International Monetary Fund reported that global data center electricity consumption in 2023 matched levels seen in entire countries such as Germany or France.

Governments have started accelerating approvals for new energy infrastructure to support AI growth. At the same time, environmental groups warn that large-scale data center expansion may clash with net-zero climate goals.

Some local communities have also opposed new data center projects over concerns about grid strain and rising electricity prices. Despite those fears, OpenAI maintains that its AI facilities will not increase local power bills.

OpenAI is not alone in facing scrutiny. Microsoft has also come under criticism for the water usage tied to its AI data center operations, highlighting the broader industry debate over how to scale artificial intelligence sustainably.