As India races to become a global data-centre hub, the question of water use may soon become one of the country’s most urgent environmental and social concerns
Newton County in the American state of Georgia and Tusiana village near Greater Noida in Uttar Pradesh may appear to have little in common — not language, not culture, not geography. Yet both places have figured in public discussions around a growing concern: the pressure that large data-centre projects can place on local water resources.
In Newton County, residents and environmental observers have raised questions about the impact of major data-centre infrastructure on community water systems. In Tusiana, villagers living near the Yotta Data Centre Park have expressed concern over declining groundwater availability and the increasing difficulty of accessing water.
These may be local experiences, but they point towards a much larger global question. As artificial intelligence expands, how will water-stressed communities be protected?
India is currently standing at the threshold of a new industrial phase driven by data centres. The country’s data-centre capacity has grown sharply over the past few years and is projected to expand further by 2030. Industry estimates suggest that India could become one of the world’s major data-centre destinations as demand rises from artificial intelligence, cloud computing, e-commerce, digital payments, streaming platforms and government digitisation.
The investment announcements are enormous. Indian conglomerates and global technology giants are planning hyperscale data-centre projects across Mumbai, Pune, Chennai, Bengaluru, Hyderabad, Noida and other cities. The official narrative presents data centres as engines of employment, digital sovereignty, foreign investment and technological leadership.
On the surface, this appears attractive. But beneath this growth lies a serious ecological and social question: where will the water come from?
Data centres are often imagined as clean, invisible infrastructure. In reality, they are large buildings filled with servers that run continuously. Every video watched online, every photo stored in the cloud, every digital payment made, and every AI query processed depends on such servers. These machines generate enormous heat. To keep them functioning, data centres require constant cooling.
Cooling requires energy. In many systems, it also requires water.
In a country like India, where temperatures are high and heatwaves are becoming more frequent, cooling becomes a major challenge. Many data centres depend on cooling systems that use water either directly or indirectly. In evaporation-based cooling, a significant share of the water used is lost and cannot be recovered. Newer technologies can reduce water use, but they are expensive and not uniformly adopted.
This is where the promise of AI begins to meet the reality of India’s water stress.
Karnataka’s Information Technology Minister Priyank Kharge reportedly told the state Assembly that every megawatt of data-centre capacity could require around 2.5 crore litres of water annually. If such estimates are applied to India’s expanding capacity, the total requirement becomes significant. A medium-sized data centre can consume large quantities of water daily, depending on its design, location and cooling system.
Globally too, the concern is rising. Several research studies and international discussions have warned that water demand from AI-driven data centres could increase sharply in the coming years. Large technology companies have started disclosing water-use figures and speaking about efficiency, recycling and replenishment. But critics argue that the sector still lacks enough transparency, especially at the project level.
The problem becomes more serious in India because many of the country’s future data-centre hubs are located in states that already face recurring water stress. Maharashtra, Karnataka, Tamil Nadu and Telangana have seen repeated droughts, falling groundwater levels and seasonal tanker dependence in many regions. Yet these same states are attracting some of the largest investments in hyperscale data centres.
Mumbai is emerging as India’s biggest data-centre hub. Pune, Bengaluru, Hyderabad, Chennai and Noida are also growing rapidly. This means the geography of India’s digital future is overlapping with the geography of water vulnerability.
To understand the seriousness of this concern, one only has to look at India’s recurring urban and rural water stress. Every summer, several major cities face anxiety over reservoir levels, while many districts depend on tankers and groundwater extraction. River basins such as the Krishna and Godavari have also witnessed periods of reduced storage and stress during dry seasons.
It is in this larger context of seasonal scarcity, falling groundwater and uneven water distribution that the expansion of hyperscale data centres must be examined.
The contradiction is stark. In one part of Maharashtra, villages in Vidarbha and Marathwada wait for water tankers during summer. In another part of the same state, the Mumbai-Pune belt is being prepared for massive data-centre investment with assured land, electricity and water access. This is not merely a technical question of infrastructure. It is a question of public priorities.
Who gets water first? A farmer? An urban slum-dweller? A village school? A hospital? Or a data centre powering global digital services?
India is already one of the world’s largest users of groundwater. In many regions, groundwater is being withdrawn faster than it can naturally recharge. This is not a temporary drought-like crisis. It is a structural crisis built over decades of over-extraction, weak regulation, poor urban planning and unequal distribution.
In such a situation, the rapid expansion of water-intensive digital infrastructure could intensify existing tensions. Water scarcity is already a source of social and political conflict in India. Inter-state disputes such as the Cauvery issue show how sensitive water-sharing questions can become. If data-centre clusters expand without water accountability, they may add another layer to such disputes.
Climate variability can further complicate the situation. Years of higher temperatures, weaker rainfall or delayed monsoons increase cooling demand and worsen water stress. Higher temperatures mean servers need more intensive cooling, while local communities simultaneously face greater pressure on drinking water and irrigation sources. This creates a vicious cycle in which climate stress and digital infrastructure demand can reinforce each other.
There are possible solutions. Closed-loop cooling systems can reduce freshwater use. Treated wastewater can be used instead of fresh water. Air-cooled and liquid-cooled systems can reduce dependence on evaporative cooling. Data centres can be located in areas with lower water stress. Renewable energy, improved server design and better heat management can reduce environmental pressure.
But these solutions require regulation, transparency and political will.
At present, India does not have a strong public disclosure framework requiring every data centre to regularly report its water consumption, cooling method, source of water, groundwater impact and recycling performance. Without such disclosure, local communities cannot know how much water is being used, from where it is being drawn, and what impact it may have on them.
This lack of transparency is one of the central problems. Some companies claim to use efficient cooling systems or treated water. Some claim to be water-positive. But unless independent monitoring is mandatory, such claims remain difficult for citizens to verify.
India urgently needs a data-centre water accountability framework. Before approving large projects, authorities should require detailed studies of groundwater levels, rainfall patterns, local water demand, existing scarcity, recharge capacity and alternative water sources. Projects in critically water-stressed zones should either be restricted or placed under strict conditions.
Local self-governing bodies must also have a role. Gram panchayats, municipalities and district administrations should not be bypassed in the name of investment speed. Public consultation must be made mandatory, especially in regions where local communities depend on groundwater.
The gender dimension of this crisis also deserves attention. In rural India, the burden of fetching water still falls disproportionately on women and girls. When groundwater levels fall, they walk longer distances, spend more time collecting water, and lose opportunities for education, health and income. Water scarcity caused or worsened by industrial demand is therefore not only an environmental issue. It deepens gender inequality.
The poor will carry the heaviest burden. Tribal farmers, landless labourers, women, and urban slum-dwellers are least responsible for the AI boom, but they may be the first to experience its water cost. A server farm may appear modern and efficient from the outside, but if its growth quietly dries nearby wells or diverts scarce public water, the development model becomes ethically questionable.
Large global technology companies have begun disclosing water-use figures for their data-centre operations, and these disclosures have intensified public scrutiny of the sector’s environmental footprint. Companies such as Amazon, Google and Microsoft have also announced measures related to water efficiency, recycling and replenishment. However, critics argue that project-level transparency remains limited, especially in countries where local communities may not have easy access to information about water sources, cooling systems and actual consumption.
Technology is never entirely neutral. Whether it becomes liberating or destructive depends on how it is governed and whose interests it serves. India is one of the world’s most ambitious technology-adopting countries. But it is also a deeply unequal society under severe environmental stress. In such a country, the expansion of AI infrastructure must be judged not only by investment figures and employment claims, but also by its impact on water, land, energy and vulnerable communities.
Some experts argue that India does not lack water, but lacks proper management, storage and distribution. There is truth in this. India receives significant rainfall, yet much of it is lost because of poor planning, inadequate storage, encroached water bodies and damaged recharge systems. But this argument cannot be used to ignore present reality. Water reforms have been pending for decades. Data-centre demand will not wait until those reforms are completed.
Some academic estimates have suggested that AI queries may carry an indirect water footprint because data centres require cooling and electricity generation also consumes water. Exact figures vary widely depending on the location of the data centre, cooling method, energy source and time of use. Still, the larger point remains important: even digital actions that appear invisible can have real environmental costs somewhere in the physical world.
Therefore, the responsible path is not to reject AI infrastructure, but to regulate it before it becomes another extractive industry.
Before tax concessions, subsidies and fast-track clearances are granted, water accountability must be made compulsory. Every large data-centre project should publicly disclose its annual water requirement, cooling technology, water source, recycling plan and local impact assessment. Use of treated wastewater should be prioritised wherever possible. Freshwater use should be strictly limited in water-stressed regions.
India must also create a national map linking data-centre approvals with water stress. No region should become a digital hub by silently transferring its water burden to farmers, women and the poor.
AI has opened new horizons of development. It can transform healthcare, education, governance, research and industry. But the country must ask a simple question: development for whom, and at whose cost?
If the dream of artificial intelligence rests on the thirst of real communities, then that dream is morally incomplete.
The crisis may not fully confront us today, but it is clearly visible on the horizon. If the right decisions are not taken now, tomorrow may arrive in the form of a dry tap.
India must decide whether its AI revolution will be built on accountability and sustainability — or whether the future of technology will be written in the language of water scarcity.
