By Titik Puspa 
Boston, MA – June 9, 2026 – OpenAI CEO Sam Altman has forcefully countered prevailing narratives surrounding the environmental impact of artificial intelligence, particularly concerning water and energy consumption, during a prominent event hosted by The Indian Express in India. While attending a significant AI summit in the country, Altman addressed what he termed "totally fake" claims about AI’s water usage, while acknowledging the historical context of such concerns and advocating for a swift transition to renewable energy sources to meet the burgeoning global demand for AI.
The controversy surrounding AI’s water footprint has been amplified by viral social media posts and online discussions. Altman specifically called out claims suggesting that a single ChatGPT query consumes a substantial amount of water, characterizing these assertions as "completely untrue, totally insane, [with] no connection to reality." He clarified that the outdated notion of AI’s significant water usage stemmed from older data center cooling methods, specifically evaporative cooling. "Now that we don’t do that," Altman explained, "you see these things on the internet where, ‘Don’t use ChatGPT, it’s 17 gallons of water for each query’ or whatever. This is completely untrue, totally insane, no connection to reality." This direct refutation aims to dispel widespread misinformation and provide a more accurate understanding of current AI infrastructure.
However, Altman was quick to pivot from dismissing water usage concerns to emphasizing the undeniable reality of AI’s energy consumption. He conceded that it is "fair" to be concerned about "the energy consumption – not per query, but in total, because the world is now using so much AI." This acknowledgment underscores the critical need for sustainable energy solutions to power the exponential growth of AI technologies worldwide. Altman’s proposed solution is clear and urgent: "we need to move towards nuclear or wind and solar very quickly." This highlights a proactive stance, recognizing that the long-term viability and ethical deployment of AI are inextricably linked to our ability to power it with clean and sustainable energy.
The lack of mandated transparency from tech companies regarding their energy and water usage has led independent researchers to embark on their own investigations. Scientists are actively attempting to quantify these environmental costs, as highlighted by NPR’s reporting on efforts to study AI’s energy and water footprints independently. This academic pursuit is crucial in a landscape where regulatory frameworks have yet to catch up with the rapid advancements in AI. Furthermore, the escalating costs associated with electricity have brought the energy demands of data centers, and by extension AI, into sharp focus, as reported by TechCrunch. These rising prices create a dual pressure – economic and environmental – on the industry to find more efficient and cost-effective solutions.
During the interview, the interviewer posed a question, referencing a prior discussion with Bill Gates, about whether a single ChatGPT query consumes the equivalent of 1.5 iPhone battery charges. Altman unequivocally dismissed this, stating, "There’s no way it’s anything close to that much." This directly challenges another frequently cited metric that contributes to the perception of AI’s extreme inefficiency.
Altman further criticized the prevalent method of comparing the energy required to train an AI model with the energy needed for a single human inference query, labeling such comparisons as "unfair." He argued that this approach distorts the true picture of AI’s efficiency. "But it also takes a lot of energy to train a human," Altman countered, drawing a parallel to the extensive developmental process of human intelligence. He elaborated on the profound energy investment in human development, spanning "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." This philosophical perspective emphasizes the immense, albeit historical and evolutionary, energy investment in creating human cognitive capacity.
From Altman’s viewpoint, a more equitable comparison would be to assess "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." This proposed metric shifts the focus to the operational efficiency of trained AI models versus the real-time energy expenditure of human cognition for similar tasks. By framing it this way, Altman suggests that AI’s energy efficiency, when measured against its functional output after the initial training investment, may be far more competitive than commonly perceived. This perspective invites a nuanced discussion about how we define and measure energy efficiency in the context of both biological and artificial intelligence.
The full interview, providing a comprehensive exploration of these topics and more, is available for viewing. The specific discussion regarding water and energy usage commences at approximately the 26-minute and 35-second mark, offering viewers direct access to Altman’s detailed explanations and arguments. The context of this discussion is the broader technological landscape and the increasing integration of AI into daily life, necessitating a clear and accurate understanding of its resource implications.
The implications of Altman’s statements extend beyond mere public relations; they signal a strategic direction for OpenAI and, by extension, the AI industry. The emphasis on renewable energy sources like nuclear, wind, and solar is not just an environmental plea but a practical necessity for scaling AI capabilities responsibly. As AI models become more complex and data processing demands increase, the energy infrastructure supporting them must be robust, sustainable, and increasingly decarbonized. The industry’s future hinges on its ability to innovate not only in algorithmic development but also in energy procurement and infrastructure.
The independent research efforts into AI’s environmental impact are crucial for establishing credible data points that can inform both public discourse and policy decisions. When official disclosures are lacking, the scientific community plays a vital role in providing an objective assessment. The current landscape, where speculation and misinformation can easily flourish, necessitates a data-driven approach to understanding AI’s true environmental footprint. The interconnectedness of rising energy prices and the operational costs of data centers further amplifies the urgency for energy-efficient AI solutions and the adoption of more affordable, cleaner energy sources.
Altman’s analogy of human development serves as a thought-provoking reminder of the long and resource-intensive process that has shaped human intelligence. By drawing this parallel, he aims to contextualize the energy investments made in training AI models, suggesting that a direct, per-query comparison without considering the evolutionary and developmental context of human intelligence might be an oversimplification. This perspective encourages a more holistic evaluation of AI’s resource utilization, moving beyond simplistic metrics to a more comprehensive understanding of its overall lifecycle impact.
The call for a swift transition to renewables is a recurring theme in discussions about the future of technology and climate change. For the AI sector, which is projected to consume a significant and growing portion of global energy, this transition is not merely an option but a imperative. The development of more energy-efficient AI algorithms, hardware, and data center designs will be critical alongside the expansion of clean energy infrastructure. Companies like OpenAI are at the forefront of this technological wave, and their leadership in advocating for sustainable practices can set a precedent for the entire industry.
The ongoing dialogue surrounding AI’s environmental impact is a critical component of its responsible development and deployment. While concerns about resource consumption are valid and should be continuously addressed, it is equally important to rely on accurate data and nuanced comparisons. Sam Altman’s recent statements offer a strong counterpoint to sensationalized claims, emphasizing the need for a more informed and balanced discussion that acknowledges both the challenges and the potential solutions for powering the future of artificial intelligence sustainably. The industry’s ability to navigate these environmental considerations will be a defining factor in its long-term success and its contribution to a sustainable global future.