Skip to main content

Thirsty servers, thirsty people: Can India build an AI future without draining its water?

 
By Atanu Roy* 
As India races to become a global digital powerhouse, a quiet crisis is unfolding beneath the surface. The same water that cools the nation's data centres is desperately needed by communities still struggling for a safe drink. The choice isn't between technology and survival—it's about getting the sequencing right.
The Fundamental Question
Safe drinking water or AI data centres? For policymakers in the developing world, this might sound like a false binary, but it is rapidly becoming one of the defining dilemmas of our time.
The two are not truly competing needs—they serve vastly different levels of development. Drinking water is a basic human necessity. Safe water directly affects survival, public health, education, and productivity. According to the World Health Organization, as recently as 2024, approximately 2.1 billion people still lacked safely managed drinking-water services, with low-income countries facing particularly severe gaps.
Unsafe water contributes to diseases including diarrhoea, cholera, dysentery, typhoid, and hepatitis A. The WHO estimates that inadequate water, sanitation, and hygiene (WASH) contributes to more than one million diarrhoeal deaths annually. Beyond the health toll, there is a staggering economic dimension: when water is not available near homes, people—especially women and children—spend substantial time collecting it, sacrificing educational opportunities and productive work.
Data centres, on the other hand, represent economic infrastructure. They support cloud computing, banking and digital payments, artificial intelligence, e-commerce, government digital services, telecommunications, and employment. A developing country does need data centres if it wants to participate fully in the digital economy.
But here is the crux: it is a question of priority and sequencing. A country with villages that lack reliable drinking water should not regard a large hyperscale data centre as a higher development priority than providing safe water to those villages.
Where the Debate Turns Controversial
The issue becomes incendiary when we examine the water consumed by data centres themselves.
Large data centres require substantial resources, including electricity and water for cooling the heated semiconductor chips in servers. Therefore, a developing country must be careful about allowing data-centre expansion in areas already experiencing water stress.
It is difficult to justify a situation where a community struggles to obtain safe drinking water while a nearby data centre receives preferential access to scarce water resources. In essence, that becomes a question of development priorities and environmental justice, rather than simply technology versus water.
This does not mean developing countries should reject data centres. Quite the opposite: they should build them strategically, using renewable energy where feasible, water-efficient cooling, recycled or non-potable water where appropriate, and locating them where they do not compete with essential community water supplies.
The big picture is clear: a data centre can accelerate economic development, but safe drinking water makes development possible in the first place.
India's Twin Ambitions
This tension is no longer theoretical. It is an increasingly urgent policy decision for countries such as India, where rapid AI and cloud expansion is occurring alongside significant regional water stress.
India has two competing ambitions. On one hand, it wants to attract massive investments in AI and cloud computing, hyperscale data centres, semiconductor and digital infrastructure, digital public services, financial technology, and data-driven industries. This is economically attractive because data centres bring investment, skilled employment, tax revenues, and supporting infrastructure.
On the other hand, India faces a fundamental water challenge: it holds about 17% of the world's population but only around 4% of its freshwater resources. Water availability is also extremely uneven geographically and seasonally.
The issue, therefore, is not: "Should India have data centres or drinking water?"
The real question is: "Where, how, and under what conditions should India allow water-intensive data centres to be built?"
The Physics of the Problem
A large data centre contains thousands of servers that generate enormous heat. That heat must be removed continuously. Some cooling systems use water. The larger the facility, the greater the potential water requirement. This creates a particularly sensitive situation if a data centre is established in a water-stressed city.
The ethical question then becomes: Should scarce freshwater be allocated to cooling servers when households don't have dependable drinking water?
There is no simple "yes" or "no," because the data centre could also generate substantial economic benefits. But the government should impose conditions.
Separating Drinking Water from Industrial Water
India should distinguish sharply between drinking water and industrial water. This distinction is particularly important.
A data centre does not necessarily have to use potable drinking water. It could potentially use treated wastewater, recycled water, rainwater, municipal sewage treated to appropriate standards, and other non-potable sources. This makes far more sense than diverting high-quality freshwater to cool servers.
Drinking water should be reserved primarily for people; treated and recycled water should increasingly be used for industry.
Location Is Everything
India shouldn't automatically put every data centre in Mumbai, Bengaluru, Chennai, Hyderabad, or Delhi simply because these cities already have strong IT ecosystems.
Government and industry should consider water availability as a major factor in site selection, alongside electricity, fibre connectivity, land, and taxation.
A data centre in a water-abundant location using recycled water could be far less problematic than an identical facility in a severely water-stressed area.
Five Principles for India's Policy
What should India's policy framework look like? Here are five guiding principles:
1. Drinking Water Gets Priority
Government should explicitly establish that essential domestic drinking-water requirements have priority over discretionary industrial consumption in water-stressed areas.
2. No Potable Water for Cooling Where Alternatives Are Feasible
Large data centres should progressively be required to use treated wastewater, recycled water, or other suitable non-potable sources for cooling.
3. Mandatory Water-Impact Assessment
Just as environmental impacts are assessed, large data-centre projects in water-stressed areas should disclose: annual water requirement, source of water, peak summer requirement, groundwater dependence, percentage of recycled water used, wastewater discharge, and impact on the local water balance.
4. Water-Neutral or Water-Positive Targets
This could become particularly interesting. A company operating a large data centre could be required to compensate for its consumptive water use by funding projects that restore or create an equivalent amount of usable water capacity—for example, wastewater recycling, watershed restoration, or aquifer recharge.
However, such schemes need rigorous measurement; simply planting trees and calling the project "water positive" would not be sufficient.
5. Don't Stop Digital Development
India cannot afford to say: "We have water problems, therefore we won't build data centres." That would sacrifice an important part of India's future economy.
Instead: Build the digital infrastructure—but make water efficiency a condition of doing business.
The Deeper Irony
For a country like India, AI and cloud computing can increase productivity and wealth, potentially helping millions of people. But water is not merely an economic resource. It is a basic human necessity.
There is a fascinating irony here: India's digital economy is built on data, but its physical infrastructure ultimately depends on very old-fashioned resources—land, electricity, and water.
That is why India's AI and data-centre expansion should be planned not merely as a technology policy, but as a water, energy, and urban-planning policy as well. Only a holistic approach can ensure that the country's digital future does not come at the cost of its people's most fundamental need.
---
*Contact: atanusingapore@gmail.com

Comments

TRENDING

Arrival of USS Lincoln: A ‘leisure trip’ to Thailand or a neo-colonial move to commodify women?

By Benyasiri Eimviriyapong   The arrival of the USS Abraham Lincoln (CVN-72) for a ‘leisure trip’ in Thailand before returning to its campaign of destruction against the people of West Asia demands immediate interrogation and opposition. For us Thai people, and those in sites of American plunder across the globe, this a rattling of the US’ cold war colonial chains on its most loyal lap dog, while conducting a war of aggression in Iran and the ongoing genocide in Palestine. In allowing the USS Lincoln to dock, the Thai ruling elites are once again facilitating a system of imperial violence that extends far beyond Southeast Asian borders, a system which in our case, is predicated on sexual violence and femicide.

Global financial institutions 'ignored' climate warnings in Nepal’s hydropower disaster: SANDRP probe

Counterview News  A catastrophic flood disaster across Nepal’s Himalayan river valleys has triggered urgent international scrutiny over the role of multilateral lenders and development finance institutions. According to an extensive investigation published by the South Asia Network on Dams, Rivers and People (SANDRP), international financiers—including the World Bank Group’s International Finance Corporation (IFC), the Asian Development Bank (ADB), the Asian Infrastructure Investment Bank (AIIB), and the Export-Import Bank of China —pushed forward major hydropower projects in disaster-prone valleys despite documented warnings, recurring historical floods, and unheeded safety recommendations.

Pilgrimage or pollution? The waste crisis of Indian faith

By Raj Kumar Sinha*  In today's times, there are certain issues that are constantly talked about and discussed, yet no visible impact seems to be made. Beyond water, soil, and pollution, one crisis that continues to be overlooked is waste. Like other life-threatening problems, waste too is gradually moving toward burying and obliterating us. After the Kanwar Yatra in Haridwar, approximately 8,000 tonnes of waste were left behind on the riverbanks, including bottles filled with urine.