UN and India on AI Data Centers: Two Divergent Views
While the UN highlights the unequal global concentration of hardware and advanced computing power among a few tech firms and nations, especially US and China, India wishes to manufacture its varied infrastructure locally to ensure that the data remains within the country’s border
There is a fundamental difference between the way United Nations talks about AI data centres and what the Indian government hopes to achieve through them.
This difference can best be captured between what the Secretary General Antonio Guterres recently said at London Climate Action Week about such data centres, and the vision that the Indian government’s Ministry of Electronics and Information Technology (MeitY) laid out in a draft policy document in 2020.
The Secretary General asked the tech firms (located mostly in US and China) to ‘come clean about the hidden environmental footprint’ of such centres. He specifically spelt out a framework for the tech firms called 'AI Environmental Transparency Initiative', and asked them to publicly measure and disclose their full environment impact, especially those related to carbon emissions, land and water usage.
In addition, he even called upon these firms to go completely Green by 2030. Guterres’ focus was obviously on using renewable energy and reducing the impact on environment.
As is clear, United Nations focus is not on the data centre per se, but the environmental risk and damage that they will cause.
That danger is real.
The 2026 report of the Canada-based United Nations University Institute for Water, Environment and Health (UN-IWEN), the academic arm of UN, spells it more vividly. Titled “Environmental Cost of AI’s Energy Use: Carbon, Water, and Land Footprints”, the report estimates that in 2030 AI’s water use will match the needs of 1.3 billion people, while its power use will be triple that of Pakistan, Bangladesh and Nigeria combined — countries with a total population of 650 million. If the data centres all over the world were treated as a nation, they would be the 11th largest one to consume electricity in 2025.
The same report, in fact, pointed out to the relationship between, what it calls, ‘carbon blindspot’ and water consumption. It said that should AI data centres transition to using Green Energy, their carbon emissions may come down by as much as 70 percent, but in so doing, their local water consumption will increase 30-fold.
What is perhaps not so well known is the requirement of critical minerals that AI demands. If the present trend continues, the demand for these minerals will grow up to four times of what we consume today. For instance, the demand for cobalt and lithium, both critical minerals, will shoot up by 500 percent by the year 2050.
It is the less developed countries that have surfeit of these minerals (countries such as Chile and Democratic Republic of Congo) and their extraction will lead to pollution, water contamination, exploitative practices, and health hazards. However, users of these critical minerals will be located in the more advanced world and would be immune to such devastations. Some experts have called this form of exploitation as a new form of imperialism.
India's view of AI data centres
The Indian government’s Ministry of Electronics and Information Technology (MeitY) has a very different view of AI data centres than UN’s. For it, the AI data centres sit at the centre of economic sovereignty and self-reliance. Located in the larger Atmanribhar Bharat policy framework, it views the data centres as protecting the ‘digital sovereignty of the country.’
While the UN highlights the unequal global concentration of hardware and advanced computing power among a few tech firms and nations, especially US and China, MeitY wishes to manufacture its varied infrastructure locally to ensure that the data remains within India’s border.
Moreover, the United Nations is concerned about resource depletion with little or no benefit trickling down to vast numbers of the local population. MeitY, on the other hand, does not refer to any of this. Instead, it factors high energy demand into national power planning, transitioning to renewable energy (50% non-fossil capacity) and nuclear energy. It plans to provide heavily subsidised IndiaAI Compute Portal access to 10,000 and more high-powered Graphics Processing Units (GPUs) for startups and researchers.
In simple terms, the Indian government has worked out a detailed four-pronged strategy where it will act as a central aggregator and subsidiser thereby helping startups and researchers from not buying expensive hardware.
Finally, MeitY actively promotes data centre capacity through subsidies and financial incentives (as enunciated in the revised policy of 2024 and 2025). The UN, on the other hand, does not refer to subsidy or accessibility but focus on regulation of AI data centres.
Yes, there is a world of difference between the UN clarion call to rein in the runaway growth of data centres and Indian government’s rush toward joining a race it doesn’t wish to lose.
[The writer, Pradeep Krishnatray, is a professor at IFHE University, Hyderabad. He was earlier the Research Director at Johns Hopkins Center for Communication Programs (JHUCCP) in New Delhi.]
Follow ummid.com WhatsApp Channel for all the latest updates.
Select Language to Translate in Urdu, Hindi, Marathi or Arabic
