Data centers are becoming the backbone of the global AI revolution. But as countries compete to build more computing infrastructure, a difficult question is emerging: how much electricity and water can local communities afford to give up for the digital economy?
Across parts of the United States, communities and governments are increasingly questioning the rapid expansion of large data centers. The concerns are not about technology itself, but about the massive physical resources required to operate modern cloud and AI infrastructure.
For India, this should be a warning—not necessarily a reason to stop building data centers, but a reason to build them much more carefully.
The hidden cost behind the AI boom
When people hear the term AI, they often think about software, algorithms and digital services. But AI ultimately depends on physical infrastructure. Large AI models require thousands of high-performance GPUs running inside data centers. These processors consume enormous amounts of electricity and generate significant heat.
That creates a chain of requirements:
AI demand → GPUs → Data centers → Electricity → Cooling → Water and infrastructure
The larger the AI industry becomes, the greater the physical infrastructure required to support it. And this is where the debate around data centers becomes much more complicated.
Why are communities in the U.S. pushing back?
The United States has experienced an extraordinary expansion of data-center projects, particularly because of the rapid growth of cloud computing and generative AI.
However, large facilities can place significant pressure on local infrastructure.
1. Electricity demand
A hyperscale or AI-focused data center can require hundreds of megawatts of power, while very large campuses can eventually require substantially more. That demand doesn’t disappear simply because the computing happens “in the cloud.” Utilities must provide the electricity, transmission infrastructure and generation capacity needed to keep these facilities operating around the clock.
In some areas, this has created concerns that existing electricity infrastructure may not be sufficient to support both data centers and ordinary residential and industrial demand.
This raises another question:
Who ultimately pays for the additional infrastructure?
If billions of dollars need to be invested in new transmission lines, substations and generation capacity, the economic benefits and infrastructure costs need to be considered together.
2. Water could become an even bigger issue
Electricity receives most of the attention, but water may become one of the most important environmental issues surrounding data centers. Computing equipment generates heat continuously. Data centers therefore require sophisticated cooling systems.
Some cooling technologies rely heavily on water, particularly systems involving evaporation. The result is that a large data center can consume substantial quantities of water depending on its cooling architecture, climate, efficiency and water-reuse practices. This becomes particularly important in regions that already experience water shortages. And this is where India’s situation deserves careful consideration.
India cannot assume that its water resources are unlimited
India is rapidly positioning itself as an important destination for cloud infrastructure and AI data centers. Several states are offering incentives to attract major investments, while technology companies are looking to expand their computing infrastructure in the country.
The economic benefits can be significant:
- New investment
- Employment
- Digital infrastructure
- Cloud services
- AI development
- Technology ecosystems
- Tax revenue
- Supporting industries
But investment figures alone don’t tell the whole story. A data center doesn’t just require land and buildings.
It requires:
Land + electricity + transmission + cooling + water + connectivity + backup systems + waste management
And these resources ultimately come from the surrounding infrastructure and environment.
The question India should ask now
India should not wait until communities begin experiencing water shortages or electricity constraints before asking whether a particular data center is appropriate for a particular location. Instead, every major project should be evaluated on its total resource footprint.
For example:
How much electricity will it consume?
Not just the initial requirement, but the expected demand after expansion.
Where will that electricity come from?
Will it depend primarily on the existing grid, or will the operator invest in dedicated renewable generation and storage?
How much freshwater will it consume?
This is perhaps one of the most important questions. A project located in a water-stressed region should not be treated in exactly the same way as one located in an area with abundant water resources.
What cooling technology will be used?
Direct-to-chip liquid cooling, closed-loop systems, immersion cooling and other technologies can potentially reduce freshwater consumption compared with more water-intensive approaches.
Will recycled water be used?
Where technically and economically feasible, treated wastewater should be considered instead of relying heavily on freshwater.
India should learn from the U.S.—not simply copy it
The lesson from the U.S. should not be:
“Data centers are bad, so India should stop building them.”
That would be an oversimplification. Data centers are essential infrastructure for the modern digital economy, and India needs more of them if it wants to become a major player in cloud computing and artificial intelligence.
The real lesson is:
Build the infrastructure—but understand its environmental and infrastructure costs before approving massive projects.
India has an opportunity that some other countries did not have: we can learn from the problems emerging elsewhere before experiencing them at the same scale.
Could water become India’s data-center bottleneck?
It is possible.
Imagine two states competing to attract a massive AI data center. State A offers cheap land and electricity but has significant water stress. State B has better access to recycled industrial or municipal water and sufficient power infrastructure. If both states offer similar economic incentives, the second location may ultimately be more sustainable. This means India’s future data-center strategy should not simply focus on:
“How much investment can we attract?”
It should also ask: “Can our infrastructure and environment support this investment for the next 20–30 years?”
The solution: make efficiency part of the approval process
India does not necessarily need fewer data centers.
It needs better data centers.
Future projects should increasingly be evaluated using metrics such as:
- Power Usage Effectiveness (PUE)
- Water Usage Effectiveness (WUE)
- Renewable-energy usage
- Freshwater consumption
- Wastewater recycling
- Cooling efficiency
- Local grid impact
- Local water availability
- Carbon emissions
- Long-term infrastructure requirements
Projects with significantly lower resource consumption should receive stronger incentives. Water-stressed regions should face stricter requirements for freshwater use and wastewater recycling. And developers should be transparent about how much electricity and water their facilities actually consume.
AI should not create a water crisis
There is an important contradiction at the heart of the AI boom. We are developing increasingly intelligent digital systems while their physical infrastructure can consume enormous amounts of electricity, land and water.
That doesn’t mean AI development should stop. It means technological progress needs to become more resource-efficient. India has an opportunity to build a model in which AI growth and environmental sustainability develop together.
If we wait until local communities are forced to choose between industrial demand and basic resources such as water and electricity, the debate will already be too late.
Conclusion
The U.S. experience should be viewed as an early warning for India. Data centers can bring enormous economic and technological benefits, but they also create significant demands on local infrastructure.
India should welcome investment—but not at any cost. Before approving a massive data center, policymakers should ask three simple questions:
Where will the electricity come from?
Where will the water come from?
And what will happen to the local community and environment if demand doubles or triples?
The AI revolution may be digital, but the resources required to power it are very physical.
India should build for the AI future—but it should make sure that future still has enough water, electricity and environmental capacity to support everyone.
