Wed, Feb 25, 2026
India’s artificial intelligence (AI) ambitions sit inside vast warehouses of servers humming along the coasts of Mumbai and Chennai, in industrial parks outside Hyderabad, and in growing clusters around Bengaluru and Noida.
As of January 2026, India hosts about 270 operational data centres, occupying nearly 23 million square metres, with installed capacity rising from 520 MW in 2020 to about 1.5 GW by mid-2025. That capacity is projected to reach 4.5–6.5 GW by 2030, backed by nearly US$95 billion in committed investments between 2019 and 2025.
The digital economy is expected to contribute 20% of India’s national income by 2030. Internet users have crossed 1 billion. AI adoption is accelerating. Data localisation mandates under the Digital Personal Data Protection (DPDP) Act and Reserve Bank of India (RBI) rules are pushing storage onshore.
The demand for data centres is not going to go down
— Dhrubabrata Ghosh Dastidar, Managing Director (Technology & Digital), Protiviti Member Firm for India
He highlighted AI, cloud adoption, and hyperscale expansion as structural drivers. For India to be “future-ready,” he argues, the country needs infrastructure first. Data centres, in that sense, are the backbone of AI sovereignty.
At the Union level, data centres have been granted infrastructure status for facilities above 5 MW, allowing developers access to easier financing and long-term credit. The Union Budget 2026 went further, announcing a long-term tax holiday extending to 2047 for data centres and cloud investments.
This strengthens India's goal of becoming a global centre for cloud, AI, and data-led services, and makes digital infrastructure a national priority. It is a significant step towards expanding the digital economy, drawing in international investment, and producing skilled jobs
— Sachin Tayal, Managing Director, Protiviti Member Firm for India
But this backbone consumes enormous power and water. Data centres alone accounted for around 0.5% of national electricity consumption and roughly 150 billion litres of water use in 2024, both projected to more than double by 2030. Cooling a 100-MW hyperscale facility can consume nearly two million litres of water per day.
The question now is whether these data centres can be built responsibly.
Electricity makes up 60%–70% of operational costs. Diesel generators remain embedded in design because uptime guarantees allow almost no margin for error. “The first thing starts from monitoring,” Dastidar said, describing telemetry systems that track water and power use in real time. Without measurement, he argues, neither efficiency nor regulation can work.
Yet monitoring alone may not resolve deeper trade-offs.
India does not yet have a binding national data centre policy. A draft released in 2020 remains unnotified. In its absence, 15 states have rolled out dedicated data centre policies or embedded incentives within IT and industrial frameworks.
States offer electricity duty exemptions, concessional land, single-window clearances, and, in some cases, renewable-linked incentives. Odisha’s 2022 policy, for instance, reimburses transmission and wheeling charges for captive renewable power. But sustainability provisions are uneven. Only five of 15 state policies explicitly embed sustainability measures: Rajasthan, Odisha, Karnataka, Tamil Nadu, and Andhra Pradesh.
India’s regulatory approach remains largely incentive-driven. Unlike the EU, which is tightening efficiency disclosures under its Energy Efficiency Directive, or certain US states that have begun scrutinising water and grid impacts after the strain seen in Northern Virginia, India has yet to impose data centre–specific environmental caps.
However, sectoral laws do apply. “The Environment Protection Act governs emissions and groundwater extraction. Data centres must obtain no-objection certificates for groundwater use. Cybersecurity rules require incident reporting within six hours and retention of ICT logs in India for 180 days,” a manager at a Noida-based data centre said.
ESG — short for environmental, social, and governance — should be more than a corporate buzzword when done right. It is increasingly a benchmark that reflects the sustainability, societal impact, and governance strength of companies, including data centre operators. Globally, investors, regulators and customers are expecting measurable performance on these three dimensions.
At its core, ESG asks three questions:
How does an operation impact natural resources and climate?
How does it affect people, i.e. employees, communities, and customers?
Is the company accountable, transparent, and well-run?
This framework helps weigh the pros and cons of mushrooming data centres.
For data centres, the E in ESG is the most visible. Power procurement, cooling technologies, water sourcing and emissions intensity are no longer just operational concerns — they are investment and regulatory risks.
The world is going to face water bankruptcy
— Nutan Zarapkar, Managing Director, Protiviti Member Firm for India
In a country where water allocation prioritises drinking needs and agriculture before industry, she warned that facilities built in stressed regions risk becoming stranded assets over time. “If you create data centres in a water-scarce area… those assets can become stranded,” she said.
Cooling choices underline the trade-offs. Water-based systems are more energy-efficient, but water-intensive. Air-cooled systems reduce water dependence but increase electricity consumption and emissions. “It is one problem while you’re trying to solve another problem,” she said, pointing to rising heatwaves and growing AI workloads.
Globally, operators are responding by tightening performance benchmarks — tracking power usage effectiveness (PUE), water usage effectiveness (WUE), investing in recycled water systems, and increasing renewable sourcing. In Europe, disclosure norms are tightening. In the US, clusters such as Northern Virginia have triggered debates over grid strain and water use.
ESG reporting, in that sense, is becoming a licence to operate.
The S (social impact) is less discussed but equally relevant. Data centres create high-skilled jobs during construction and operations, and can enable digital inclusion by powering healthcare, education and financial services platforms. But they also draw heavily on local land, power, and water infrastructure. Smaller facilities, in particular, often rely on municipal water supply. That can create friction in already stretched cities.
The G (governance) may ultimately determine credibility. Zarapkar raised concerns about whether boards and leadership teams are adequately equipped to assess climate and ESG risks. She pointed to the idea of “double materiality”, not just how environmental risks affect a company’s finances, but how the company’s operations affect ecosystems and communities.
“The first thing starts from monitoring,” Dastidar said, underscoring that credible data on water and power use is essential if ESG is to move beyond reporting into real risk management. He also maintained that stronger implementation may matter more than new rules. “You don’t have to bring in new regulations. The principles already exist,” he said.
That puts the spotlight on execution. “The persistent focus on ESG principles underscores the belief that contemporary growth must be grounded in transparency, trust, and robust governance,” Tayal told The Secretariat.
The opportunity, then, is structural. India’s data centre sector is still scaling, which means renewable integration, recycled water systems, storage-backed grids, and clearer ESG disclosures can be embedded early, rather than retrofitted later.
“Companies that will combine growth with resilience and strong governance will be best equipped to compete on a global scale and make a significant contribution to establishing India's long-term economic leadership,” Tayal added.
The race to build capacity is on. The opportunity now is to ensure resilience scales with it.