Openomics 2026: The Risky Romance with Highscalers

 MONTHS BEFORE the announcement in this year’s Union Budget of a long tax holiday for global tech giants and others who use data cen­tres built in the country, multiple corporations, local and overseas, had announced their ambitious plans to set up such facilities, almost by force of habit.

Tata Consultancy Services announced late last year that it would tie up with private equity firm TPG to develop AI data centres. The joint venture was named Hy­perVault AI Data Centre. Similarly, Adani Enterprises’ subsidiary AdaniConneX and Google agreed to develop a massive data centre in Visakhapatnam, Andhra Pradesh, and vowed to spend billions of dollars from 2026 to 2030. Much earlier, there were reports in a section of the me­dia that Sam Altman-led AI giant OpenAI would set up a data centre in India, its sec­ond-largest market after the US and where it also has the largest base of student us­ers, aligning itself with the government’s IndiaAI Mission worth $1.2 billion that envisages making home-grown Large Language Models (LLMs) to boost na­tional AI capabilities. A Bloomberg report had said that a new data centre could be significant for OpenAI’s Stargate-branded AI infrastructure push. Outside the US and besides India, OpenAI was looking for collaborations in Abu Dhabi, Norway and other countries. SoftBank, MGX, Oracle, and OpenAI itself are all involved in funding Stargate, and OpenAI is work­ing with the US government to build an AI infrastructure to counter any threat from China, according to earlier reports (‘ChatGPT Comes to the Classroom’, Open, September 5, 2025). Microsoft’s Indian-origin CEO Satya Nadella said a few months ago that he had discussed the company’s investment plan in data cen­tres with Prime Minister Modi. Nadella also announced an investment of $17.5 billion by 2030 to help build AI infrastruc­ture in India. More such announcements kept coming out ahead of Finance Min­ister Nirmala Sitharaman’s February 1 Budget speech.

Behind such moves, clearly, was a busi­ness opportunity to correct a structural problem thanks to India being a large consumer of global data.

Which was why Debasish Mishra, chief growth officer, Deloitte South Asia, immediately lauded the governmental move in the Union Budget that was appar­ently aimed at warding off any fears of tax liabilities among global IT giants. Writ­ing on his LinkedIn page on Budget day, he argued that India currently consumes around 20 per cent of global data but hosts only around 3 per cent of global data-centre capacity. He asserted that this “structural mismatch” means that demand generated in India is being served by infrastructure and energy outside its borders.

Mishra averred that, therefore, he found immense logic in a tax holiday until 2047 for foreign companies provid­ing cloud services to customers globally, provided they use data-centre services based in India, as well as in the announce­ment of a 15 per cent safe harbour on costs where data-centre services are provided from India by related entities.

“These measures are designed to lo­calise compute, build sovereign digital infrastructure at scale and, by extension, reduce global energy distortions. Ac­cording to our assessment at Deloitte, the data-centre demand by 2030 will be over 10GW, needing an additional 45–50 million square feet of real estate space and 40–45 Terawatt Hours (TWH) of in­cremental power by 2030,” Mishra said.

NOW, WHAT EXACTLY are data centres?

According to IBM, it is “a physi­cal room, building or facility that houses IT infrastructure for building, running and delivering applications and services. It also stores and manages the data asso­ciated with those applications and ser­vices”. It also highlights the evolution of these entities: “Data centres started out as privately owned, tightly controlled on-premises facilities housing traditional IT infrastructure for the exclusive use of one company. Recently, they’ve evolved into remote facilities or networks of facilities owned by cloud service providers (CSPs). These CSP data centres house virtualised IT infrastructure for the shared use of multiple companies and customers.”

And the business potential is huge, too. A recent McKinsey analysis fore­casts that by 2030, data centres will need $6.7 trillion of worldwide investment to keep pace with the demand for compute power, around 70 per cent of which will come from AI workloads.

Economists and technology analysts that Open spoke to are excited about the intent and cautiously optimistic about the road ahead. For his part, Pravin Krishna, Chung Ju Yung Distinguished Professor of International Economics and Business at the School of Advanced International Studies at Johns Hopkins University, tells Open: “The policy smartly recognises AI infrastructure as an extraordinarily im­portant strategic asset. Success will, of course, hinge on execution and whether complementary investments in power, connectivity, and skills follow.”

For her part, Lekha Chakraborty, pro-fessor, National Institute of Public Finance, and member of Board of Management, International Institute of Public Finance, Munich, says that the 2026–27 Budget grants a 21-year tax holiday to foreign cloud providers routing global services through Indian data centres — a strategic move to establish India as a leading AI and digital infrastructure hub. “By removing tax uncertainties on overseas income, it aims to attract massive FDI from hyper­scalers,” she adds, emphasising that risks abound. “Data centres’ high power and water consumption demand stringent green mandates for such investments. The lengthy exemption entails signifi­cant tax expenditure, complicating fiscal consolidation. Alignment with Pillar 2’s 15% global minimum tax requires careful structuring to prevent top-up taxes abroad. Taxation alone is insufficient; sustained ease-of-doing-business reforms are equally critical to secure long-term AI inflows.”

Indeed, the real estate space will also see a massive churning. According to JLL Research, which analyses real estate trends and other industry insights, nearly 100 GW of new data centres will be added between 2026 and 2030, doubling global capacity. The global data-centre sector will likely expand at a 14 per cent CAGR through 2030, which will require energy innovations to alleviate grid constraints, it adds, noting that hyperscalers will remain a key driver of sector growth, executing a dual strategy of leasing and self-building. Its study forecasts that tenants will likely spend an additional $1 to $2 trillion to fit out their space with IT equipment.

Meanwhile, Bengaluru-based Sreejith Sreedharan, a technology analyst, author and Aula Fellow in AI Science, Technology and Policy, points out that the Indian government’s 21-year tax holiday for overseas companies using large data centres in India can attract investment, but it calls for cautious optimism for two reasons: the environmental strain on a water- and power-constrained India, and the rapid pace of technological change that shortens asset life cycles.

Data centres consume large amounts of electricity and water, risks that could further stress grids and supplies in states such as Maharashtra and Gujarat unless the incentive is conditional on strict sus­tainability requirements, adds Sreedharan, whose latest book is Future of Work – AI Augmented, Autonomous, Decentralised. “A shorter, reviewable horizon of about 10 to 12 years with built-in policy reviews and conditional renewals would better balance near-term gains with long-term environ­mental and fiscal responsibilities. This is true for all developing countries,” he says.

According to him, the 21-year lock-in also assumes a stable geopolitical and technological future. A fall in global de­mand caused by trade frictions, shifting data-localisation rules, or reduced cloud reliance could leave large facilities unde­rused, and emerging technologies such as edge computing, quantum processors, or decentralised AI may change where and how computing happens, reducing the value of today’s centralised builds. “A shorter, conditional incentive window would better protect public finances and allow policy to adapt as markets and tech­nology evolve. Surely, big-tech are smil­ing,” Sreedharan argues.

True, notwithstanding all the excite­ment, there are those sceptics who put the spotlight on some real issues.

After all, the AI boom was always ex­pected to come with a huge price, on the environment. Data centres, which store and process data, text, images and other information, can wreak havoc too, as news reports all the way from Georgia in the US to small towns in Europe confirm. Data centres may power everything from online bank transactions to AGI tools like ChatGPT, but also cause water and power shortages.

A report by the UN Environment Pro­gramme (UNEP) had flagged this two years ago. It said, “Most large-scale AI deployments are housed in data centres, including those operated by cloud service providers. These data centres can take a heavy toll on the planet. The electronics they house rely on a staggering amount of grist: making a 2 kg computer requires 800 kg of raw materials. The microchips that power AI need rare earth elements, which are often mined in environmen­tally destructive ways… The second problem is that data centres produce electronic waste, which often contains hazardous substances, like mercury and lead. Third, data centres use water dur­ing construction and, once operational, to cool electrical components. Globally, AI-related infrastructure may soon con­sume six times more water than Den­mark, a country of 6 million, according to one estimate. That is a problem when a quarter of humanity already lacks access to clean water and sanitation. Finally, to power their complex electronics, data centres that host AI technology need a lot of energy, which in most places still comes from the burning of fossil fuels, producing planet-warming greenhouse gases. A request made through ChatGPT, an AI-based virtual assistant, consumes 10 times the electricity of a Google Search, re­ported the International Energy Agency.” Typically, some data centres consume as much as 500,000 gallons of water to cool the buildings and the computers inside.

In that sense, as India sees huge op­portunities in building data centres, the challenges are humongous, too. Clearly, a judicious use of resources is central to any success on this front.