598
Artificial intelligence is transforming the way people work, communicate and access information in this digital world. Whether you want to use it to generate images, answer questions, write emails or power business operations, AI has become a part of our daily life.
However, behind every AI prompt, there is a need for a vast network of powerful data centres which require a huge amount of electricity, computing power and water to operate. Most of the discussion around AI focuses on energy consumption and carbon emissions, but experts believe that there is another environmental challenge that deserves equal attention, if not more, i.e., water.
India is expanding its AI infrastructure and positioning itself as a global technology hub; however, there are growing concerns about how much water these facilities consume and whether the country’s water resources can keep pace.
India’s AI Data Centre Boom Comes With a Hidden Cost
India’s digital economy is expanding at an unprecedented rate, with government and industry investing heavily in AI infrastructure. According to government figures, the country’s data centre capacity has grown from around 375 megawatts (MW) in 2020 to around 1,500 MW in 2025. According to industry experts, this number could reach between 8,000 MW and 10,000 MW by 2030.
Several global technology companies are also working to expand their infrastructure in India. Google is building a major AI data centre near Visakhapatnam, while Microsoft continues expanding its cloud and AI infrastructure across multiple Indian cities. Amazon is also increasing its capacity, and seeing the upward AI trend, Reliance Industries has partnered with Meta to develop its first AI data centre in India. This growing expansion shows India’s ambition to become a leading AI destination.
Why Do AI Data Centres Use So Much Water?
Cooling Thousands of Servers Around the Clock
Every AI request is processed inside a data centre packed with thousands, and sometimes hundreds of thousands, of computer servers. These servers run complex calculations within seconds, and resultantly, it generates a significant amount of heat as they operate. To avoid overheating, these data centres need cooling systems that run continuously.
Many such AI firms use evaporative cooling, which absorbs heat before being released into the atmosphere. Since much of this water evaporates during the process, it cannot be recovered for reuse. As AI models become larger and more advanced, the amount of computing power and cooling required also increases.
AI Training and Daily Use Both Consume Resources
Training large AI models is particularly resource-intensive. According to a report by the United Nations University Institute for Water, Environment and Health (UNU-INWEH), training ChatGPT-4 is estimated to have required around 592 million litres of water.
However, experts note that the long-term environmental impact extends well beyond training. Billions of AI interactions every day continue to consume electricity and cooling resources long after a model has been deployed.
AI Data Centre Boom: How Much Water Do AI Data Centres Consume?
According to a 2026 report by the Council on Energy, Environment and Water (CEEW), a typical 100 MW hyperscale data centre can consume around 20 lakh litres of water every day for cooling, although actual usage depends on the cooling technology and local climate.
Some estimates suggest a 1 MW data centre may require approximately 68,500 litres of water daily. India’s data centres are estimated to have consumed nearly 150 billion litres of water during 2024-25. As digital infrastructure continues to expand, annual consumption could rise to around 358 billion litres by 2030.
The shift towards larger AI-focused facilities could further increase demand for both electricity and water in the coming years.
Why Experts Are Raising Concerns
The challenge is not only how much water data centres consume, but also where they are located. According to CEEW, more than 65 per cent of India’s existing data centre capacity is concentrated in Mumbai, Chennai, Hyderabad, Bengaluru and Noida.
Most of these cities where these data centres are located are already facing varying levels of water stress. Bengaluru’s water crisis in 2024 highlighted how rapidly urban water supplies can come under pressure when groundwater extraction exceeds natural recharge. Chennai has previously faced a near “Day Zero” situation, while Hyderabad is expected to experience increasing water shortages in the future.
Climate Change Could Intensify the Challenge
Higher temperatures increase the amount of cooling required to keep servers operating safely. At the same time, changing rainfall patterns and prolonged dry spells may reduce the reliability of freshwater supplies.
The UNU-INWEH estimates that global data centres were associated with around 4.5 trillion litres of water consumption through electricity generation in 2025. If current trends continue, that figure could exceed 9 trillion litres annually by 2030.
AI’s Environmental Impact Goes Beyond Water
Large data centres also require substantial amounts of electricity, occupy large parcels of land and generate electronic waste over time. Experts also point to concerns such as heat island effects and noise pollution around major facilities. As AI infrastructure grows, balancing technological progress with environmental sustainability will become increasingly important.
Can AI Data Centres Reduce Their Water Footprint?
Experts think that these issues can be controlled if sustainable methods are adopted early. Using recovered or treated wastewater for cooling AI systems is one of the most promising ways to reduce dependency on freshwater.
Seawater-based cooling, hybrid dry cooling, and closed-loop liquid cooling are examples of emerging technologies that can lower water usage without sacrificing operational efficiency.
Apart from all this, selecting a spot carefully is equally important. Future data centres could lessen the strain on areas presently experiencing water scarcity by being situated in areas with more reliable water availability.
India Wants to Become an AI Superpower: The Road Ahead
Over the next decade, India’s aspirations for AI are anticipated to drive unprecedented expansion in digital infrastructure. That expansion will support innovation, economic growth and new technological capabilities.
However, the environmental costs cannot be ignored. The question of whether AI uses water is no longer crucial. Rather, policymakers, businesses and technology companies must think in a direction to understand how to satisfy the nation’s expanding computing demand while safeguarding one of its most precious natural resources.
If India can combine AI innovation with smarter cooling technologies, sustainable water management and better infrastructure planning, it could build a digital future that is both technologically advanced and environmentally responsible.
Sudeep Singh Rawat is a Copy Editor at NewsX Digital with over five years of experience in digital journalism. He covers national and international news, with a focus on politics, crime, technology, public affairs, and trending topics. Before joining NewsX Digital, he worked with Business Standard, The Sunday Guardian, and other digital media organisations. His reporting focuses on breaking news, enterprise stories, and explainers that inform readers through accurate, balanced, and context-driven journalism.
