When Tech Billionaires Come Marching In…

The establishment of hyperscale AI data hubs have resulted in protests in Europe --Ireland, Scotland, the Netherlands, Germany, Spain, Italy, and France. And it has not been a smooth ride for technology companies in Chile, Uruguay, Brazil, Paraguay, and Mexico. And now, they are coming in big in India.

By Ajith Pillai in Chennai

Budget 2026 included a significant corporate concession that received little media attention: an extraordinary 21-year tax holiday, lasting until 2047, for foreign technology companies that invest in data centres in India. Perhaps this act of magnanimity by the Union finance minister was played down by mainstream media because the Budget was sandwiched between two hyped events —the "mother” of all trade agreements with the European Union and the "father" of all post-tariff deals with the United States.

However, in the tech world, especially among the Big Four AI companies —Google, Amazon, Microsoft and Nvidia — the announcement of a tax holiday was met with great enthusiasm. The big tech bros have already packed their bags to be in Delhi for the AI Summit starting February 16 and booked Rs 15-30 lakh per day suites at five-star hotels in Delhi.

Lesser tech mortals opted for rooms priced between Rs 3-4 lakh per day (tax not included). Apparently, the hotels strategically upped their rates for the occasion despite Trump’s zero-tariff proclamation for everything American!

At the Summit, hyper-scale generative AI-compatible data centres will undoubtedly be a major topic of discussion. Silicon Valley billionaires view India as the next port of call for establishing large data centres, thanks to an extremely supportive government in Delhi, favourable regulations, cheaper power, lower operating costs, and minimal anticipated public protests.

In stark contrast, in the United States, there has been increasing community resistance to large hyper-scale data centres that support AI innovation and training.

At the core of the objections in the US, Europe and Latin America are environmental concerns. Even tech companies acknowledge that data centres are a considerable drain on local resources, require large amounts of power from the grid, and guzzle water, which in turn depresses the water table in and around the areas where they operate. This affects the water supply for residents and farmers, who will face acute water shortages.    

The protests in the US have been widespread. A recent report in The Guardian, quoting the research group Data Center Watch, notes that $98 billion worth of projects were delayed or cancelled between March and June 2025 in the United States because of opposition from local communities. More than 20 groups across 17 states successfully opposed these projects, forcing local authorities to review the permits granted.

In Europe, protests are ongoing in Ireland, Scotland, the Netherlands, Germany, Spain, Italy, and France. And it has not been a smooth ride for technology companies in Chile, Uruguay, Brazil, Paraguay, and Mexico.

The upside of data centre proliferation in India is significant dollar inflow into the country. But what about employment? AI-compatible hyperscale data centers, which typically span over 10,000 square feet and can reach millions of square feet in total area, will create temporary jobs for thousands during the construction phase. However, once fully operational, they may employ only 200 permanent staff members per centre or even less.

Technology analyst Christopher Tozzi, writing in Data Centre Knowledge, a platform for daily news and in-depth analysis for the data centre industry notes: “While data centres may create many jobs in the short term, most result in relatively few permanent jobs – hence why critics tend to highlight the generous tax abatements  that data centres often receive. They tend to argue that local communities receive few economic benefits from becoming home to a new data centre.”

So, what are these data centres?

Simply put, they are like warehouses where data is stored in an easily accessible form. A typical data centre is packed with racks of servers, storage devices, and network equipment. It plays a critical role in processing, storing, computing, and distributing large amounts of data that support applications, cloud computing, and web-based platforms and systems.

Think social media, WhatsApp, sharing of videos or photographs, video streaming, online banking/fintech, credit card processing, UPI transactions, e-commerce, cloud storage, ride-hailing services, Google Search, Google Maps, and email, and you will have to factor in data centres. Any app or service that needs instant or remote access to data requires a data centre.

For the past 15 years, the storage and use of data have been crucial. However, data centres were often considered "hidden infrastructure," typically located in small, unobtrusive spaces. Aside from the larger facilities, most people have been unaware of their existence. Recently, with the rise of generative AI—technology that can create images, text, videos, and other media in response to user prompts—the situation has changed dramatically.

The new focus on generative AI has necessitated the expansion of, or the setting up of mega data centres. Generative AI requires substantial computational resources, which determine how quickly a system can perform tasks, ranging from local device processing to large-scale AI training. The data centre must have super-efficient hardware components to be effective.

A data centre that is compatible with generative AI requires significant amount of power to operate. Additionally, it needs large quantities of clean water to cool its systems. This highlights the crux of the problem.

According to S&P Global Commodity Insights estimates, India is expected to become the second-largest market for data centre electricity demand in the Asia-Pacific over the next two years, surpassing Japan and Australia. According to this report, India’s data centre power demand is expected to grow almost five-fold from its current level by 2030, with an annual average growth rate of approximately 28 per cent. At present, India has about 150 operational data centres. This figure could touch 800 or more in the next few years.

India’s power grid, which still relies heavily on fossil fuels, will be strained by the emergence of hyperscale data centres, jeopardising the country’s green objectives. Additionally, the renewable energy generated is not being effectively integrated into the grid due to inadequate infrastructure. Clearly, transmission is an issue.

Current discussions suggest that battery storage for solar and wind energy could be a viable solution. However, little progress has been made on this front.

Of even greater concern is the water consumption of hyperscale data centres, which are necessary for serving and training generative AI. A BBC report from last November projected that India's data centre water consumption will rise from 150 billion litres in 2025 to 258 billion litres by 2030. Furthermore, a study by S&P Global predicts that by 2030, 60 to 80 per cent of India's data centres will experience high water stress.

A field report by Down To Earth, the environment magazine supported by the Centre for Science and Environment (CSE), reveals the water stress experienced by the local population in the vicinity of a village near Noida in Uttar Pradesh and in a locality in Bangalore. The four-part report also showed that there is very little transparency among the companies that run data centres regarding water-use efficiency and water-recycling projects.

Crucially, a World Bank study notes that India has 18 per cent of the world's population, but only 4 per cent of its water resources.

Do we require data centres?

We do, but perhaps the kind that serve our interests. India is projected to have the highest data consumption globally by 2028. This surge is driven by the increasing number of internet and mobile users. According to the Internet and Mobile Association of India (IAMAI), there are currently 958 million active internet users in the country, with 57 per cent (approximately 548 million) located in rural areas. Additionally, AI chatbots like ChatGPT have their second-largest user base in India.

While the government welcomes tech billionaires, it must ensure that their data centres comply with regulations and do not adversely affect the livelihood and environment of local communities. Last year, Google’s $15 billion investment in an AI data centre in Visakhapatnam, Andhra Pradesh, raised concerns.

Environmental groups have raised objections, noting that Visakhapatnam is already facing an acute water crisis and that the data centre will exacerbate the problem.

As for generative AI, it’s still early days. An MIT report titled The GenAI Divide: State of AI in Business 2025 reveals that US businesses invested between $35 billion and $40 billion in AI initiatives. However, almost all of them are experiencing little to no return on their investments or any measurable impact on profits. The operational costs and research and development expenditures appear to be too high for businesses to see financial benefit.

AI, as they say, is indeed a developing story…

---

Ajith Pillai is member, Editorial College, senior editor and writer, independentink.in.

A seasoned journalist working in the profession for 40 years, he has reported out of Delhi, Mumbai, Chennai, Andhra Pradesh and Kashmir on a broad spectrum of events related to politics, crime, conflict and social change. He has worked with leading publications, including The Sunday Observer,Indian Post, Pioneer, The Week and India Today, where he headed the Chennai bureau. He was part of the team led Editor Vinod Mehta that launched Outlook magazine and headed its current affairs section till 2012. Under his watch, Outlook broke several stories that attracted national attention and questioned the government of the day. He has written two books—'Off the Record: Untold Stories from a Reporter’s Diary,’ and a novel, ‘Junkland Journeys’. He is currently working on ‘Obedient Editor’, a satirical novel on the life and times of a ‘compromised’ journalist. The short story presented here is from a collection that is awaiting publication.