ChatGPT's Water Footprint: Fact or Fiction Behind Viral Claim

OpenAI CEO Dismisses Claims of Excessive Water Usage in AI Training

OpenAI CEO Sam Altman has disputed claims that its AI model, ChatGPT, requires large amounts of water to train. Speaking at the India AI Impact Summit 2026, Altman argued that such claims are "completely untrue" and have no basis in reality.

The debate surrounding the resource demands of artificial intelligence has gained significant attention in recent years. Data centres traditionally use substantial amounts of water to prevent overheating and cool down electrical components. According to a report by Xylem and Global Water Intelligence, water usage for cooling is expected to more than triple in the next 25 years due to the rising demand for AI compute.

However, Altman shifted the focus towards the energy consumption of AI, warning that the exponential use of AI demands a swift transition to renewable energy sources. He suggested that the world should move towards nuclear, wind, and solar energy to mitigate the environmental impact of AI.

Read also: Treasury Yields Experience Largest Increase in Two Weeks Following Release of Labor Market Data

In a separate argument, Altman compared the energy usage of AI to that of humans. He claimed that it takes approximately 20 years of life and energy to train a human being, and that AI has likely caught up in terms of energy efficiency once its model is trained. Altman's comments sparked criticism on the internet, with some netizens bashing him for his opinion.

Key Figures:

• Sam Altman: CEO of OpenAI
• ChatGPT: AI model developed by OpenAI
• Sridhar Vembu: Founder of Zoho