Author: Ayushi Raj
College: CMP Degree College, University of Allahabad
Abstract
The operation of Artificial Intelligence systems relies heavily on data center Infrastructure, which requires significant quantities of water for cooling, raising important Environmental and Sustainability concerns. In a recent article, OpenAI CEO Sam Altman mentioned that ChatGPT uses 0.3 millilitres of water per query. The official numbers, published by him in 2025, state that electricity usage is around 0.34 Watt-Hours per query, and water usage is approximately 0.000085 gallons, which is about 0.32 millilitres per query.
An American multinational Investment Banker, Morgan Stanley, has warned that AI could consume 1 trillion litres of water every year, by 2028. This Article critically examines this claim, which represents an eleven-fold increase from 2024 levels, driven largely by system cooling, power generation, and chip manufacturing. The report identifies three main sources of AI’s water usage. The first is, direct water use for cooling systems to prevent servers from overheating. The second is, involves indirect water consumption linked to electricity generation that powers data centers. The third is, relates to water usage in the production of semiconductors and chips that enable AI computing.
To the Point
We know AI performs trillions of calculations every day. But where do these calculations happen? They take place on high-end GPUs (Graphics Processing Units) and TPUs (Tensor Processing Units) stacked together, creating heat equivalent to that of three houses. How is this heat managed? Through a process known as Wet Cooling, where Water is used to evaporate and cool down the data centers. What kind of Water is this? It is our fresh water. Sam Altman stated that a single query uses 0.3 millilitres of water, which is about 1/15 of a tablespoon. However, upon reviewing some research papers and performing my own calculations, I found that this figure only accounts for water used in cooling, not for the Electricity Consumption related to the manufacturing of AI hardware such as GPUs and TPUs, nor the continuous training of AI models. So, the real question is What is the real consumption of these AI Tools? ChatGPT uses 148 million litres of water every day, which is equivalent to the water consumed by 6 million UK households. Importantly, all of this is our fresh water. We are also living in water stress zones, some places like Texas, Arizona, Georgia. They are already going through ground water depletion because of these data centers. In regions such as Lowa, Agricultural activities have reportedly been curtailed due to increasing water demands from data centers used for cooling. This raises a broader policy question: should scarce water resources be allocated to support AI applications, including relatively non-essential uses such as homework assistance and entertainment content creation? Or should we use that water for Agriculture to ensure we have Food to Eat? AI is not only costly for Companies and Employees; it is costly for the Environment as well. In the end, we rely on AI to do our work, but we may not have enough Water to Drink or Food to Eat.
Use of Legal Jargon
Legal & Environmental Issues: Although the Constitution of India does not explicitly guarantee a right to clean water, the judiciary has recognised it as an integral component of the Fundamental Right to Life under Article 21. Through various judicial decisions, the Supreme Court and High Courts have expanded the interpretation of Article 21 to include rights necessary for a dignified life, including access to clean and safe water.
Public Interest Litigation (PIL): When the right to clean water is violated, individuals can seek legal remedies by filing Public Interest Litigations (PILs) in the Supreme Court under Article 32 or High Courts under Article 226. The Judiciary has been proactive in addressing PILs related to water pollution, degradation of water bodies, and the failure of public authorities to provide safe drinking water.
Principles & Doctrines: AI development should be balanced with the following principles: –
- Precautionary Principle: When there is a risk of serious environmental harm, preventive actions should be taken even without full scientific certainty.
- Public Trust Doctrine: Natural resources like water, air, and forests are held by the state in trust for the public and must be protected for public use and benefit.
- Sustainable Development Principle: Development should meet current needs without compromising future generations’ ability to meet their own needs.
Duty of Care & Fiduciary Duties: The use of AI in water governance may impose legal obligations on water utilities and public authorities to ensure that AI-driven decisions do not harm citizens or environmental resources. Under the Public Trust Doctrine, a Fiduciary Duty may be assumed, requiring the management of water resources to prioritize public welfare over commercial interests.
The Proof
The findings emphasize that more than half of the world’s major data center hubs are located in regions already facing moderate to high levels of water stress. These include areas vulnerable to droughts, water quality degradation and declining groundwater supplies. The projected increase in water consumption could exacerbate existing water shortages, particularly in regions where competition for water between industries and local communities is already significant. The expansion of AI models, which require intensive computing power, is expected to accelerate demand for data center capacity. As industries such as Finance, E-Commerce, Healthcare, and Education increasingly adopt AI technologies, the supporting Infrastructure will require substantially greater computational capacity and cooling resources. Consequently, water consumption is expected to rise significantly, particularly in data center cooling systems, power generation facilities, and semiconductor manufacturing plants. The report notes that water-efficient cooling methods, recycling systems and a shift towards renewable energy sources could mitigate the projected rise in consumption. Adoption of advanced liquid-cooling technologies, regional regulations and corporate sustainability policies are among the factors that will determine the scale of the increase. As we all know, India, in collaboration with Israeli institutions and technology partners, has explored the use of artificial intelligence, digital twin technologies, and advanced monitoring systems to improve water distribution efficiency and minimize water losses. Furthermore, the two countries have supported joint research and innovation projects aimed at enhancing urban water management through AI-driven solutions.
However, without significant changes in industry practices, the study warns that water use by AI-related infrastructure could become a major Environmental concern. The findings suggest that careful siting of data centers and improved efficiency standards will play a critical role in balancing industrial growth with water sustainability.
Case Laws
- M.C. Mehta v Kamalnath (1997)
This important court decision reminded us that the welfare of the people is the most important law. It showed that the government has a responsibility to protect public health by making sure people have access to clean and safe water and by preventing dangers to the environment.
- Bandhua Mukti Morcha v Union of India (1984)
In this case, the Supreme Court said that bonded labourers have a right to basic things like clean water, shelter, and medical care. The Court also said that giving clean drinking water is a basic duty of the government under Article 21 of the Constitution.
- A.P. Pollution Control Board v Prof. M.V. Nayudu (1999)
The Court talked about the importance of clean water and the idea that we should be careful when dealing with the environment. It said the government must control water resources and make sure the right to clean water is not lost.
- Coffin v Left Hand Ditch Co (1882)
This case focused on how water ownership works for farming. It said that if a farmer spends a lot of effort to bring water from one area to another, they should be allowed to use it for their land. It would be unfair to take that away just because they are using a different stream.
Conclusion
The right to clean water is an essential part of the right to life as stated in Article 21 of the Indian Constitution. Through important court decisions like M.C. Mehta, the courts have recognized clean and safe water as a basic right. Laws, such as the Water Act and the
Environment Protection Act support this right, but the government has the main job of protecting and managing water for all people. At the same time, artificial intelligence (AI) can help with managing and saving water. But AI technology, especially data centres and chip manufacturing, uses a lot of water. So, as AI grows, it’s important to make sure that progress doesn’t harm water sustainability. We need to find a balance between using AI and keeping water resources safe for the future.
FAQs
Q1. Does the expansion of AI infrastructure threaten access to water, especially in drought-prone areas?
AI has both good and bad effects on water. It can help with managing water by finding leaks, predicting floods, understanding demand, and improving farming with smart irrigation. But the data centres and cooling systems that run AI use a lot of water, putting more pressure on limited supplies. This creates a big challenge for policymakers: how to use AI to improve water management without making the environment worse.
Q2. Are there any Indian laws that control how much water AI data centers use?
Right now, there are no special laws in India for controlling water use by AI data centres. But there are existing rules like the Environment (Protection) Act, 1986, the Water (Prevention and Control of Pollution) Act, 1974, and state rules about using groundwater. These can be used to manage how much water data centres take, use, and release.
Q3. How does international law look at Water Sustainability with AI growth?
International agreements like the UN Sustainable Development Goals (SDG 6 – Clean Water and Sanitation) and the Paris Agreement promote using technology in a sustainable way. There’s no special AI-water treaty, but many countries are working on rules that require companies to report how much water their AI uses.
Q4. Do AI-driven Water management systems cause Data privacy issues?
Yes, These systems collect a lot of information about how much water people use, where they live, and their habits. According to the Digital Personal Data Protection Act, 2023, such data must be handled properly and only with proper permission and security to prevent misuse or leaks.
