India recently announced a two-decade tax break for foreign tech companies that use data centers in the South Asian country to provide services to overseas customers. The exemption, though, doesn't apply to services offered to Indian clients, which will continue to be taxed. 

The move is part of India's efforts to draw investors into its fast-growing digital infrastructure sector and become a global hub for artificial intelligence (AI) and cloud computing infrastructure.

Prime Minister Narendra Modi's administration has already rolled out a slew of incentives such as granting infrastructure status for data centers to accelerate their development. Some state governments have also eased land-use rules.

New Delhi hosted a global AI summit last month, gathering those shaping the future of the technology, such as OpenAI's Sam Altman, Google's Sundar Pichai, and Dario Amodei of Anthropic. The event showed India wants its voice heard about the impacts of AI on developing countries and on whose terms AI is built and deployed.

Ambitious AI investment plans

The Asia-Pacific region is projected to attract about $800 billion (€685.4 billion) in data center investments by 2030, according to the consulting firm Deloitte.

New Delhi wants to attract a significant slice of these capital flows. Indian conglomerates like Reliance Industries Ltd., led by Asia's richest person Mukesh Ambani, Adani Enterprises and the Tata Group have also made ambitious multibillion-dollar pledges to build out the nation's AI-related infrastructure — some in partnership with US tech giants.

The investments promise to position India as a global leader in AI and cloud computing and create thousands of new jobs.

But Apar Gupta, founding director of the Internet Freedom Foundation, said that "data centers are not the same as AI leadership."

"They add servers and storage, not the ability to build and control advanced AI," he added.

Gupta's concern is that India is overly focused on infrastructure building while neglecting other key areas. 

Without investment in research, people and Indian datasets, he warned, the country risks "mainly hosting infrastructure for global firms, not shaping the technology, or setting the rules."

Divij Joshi, research fellow at the London-based Overseas Development Institute, a think tank, echoed this view.

"Hosting servers does not mean controlling what runs on them," he told DW.

"For India to gain real leverage, it would need infrastructure at a scale that makes global AI supply chains genuinely dependent on it. Given how far ahead the US and China already are, that is a long shot."

To compete with others in the global AI race, Joshi said, "India needs efficient grids, regulations, tech transfer, and R&D investment first."

Power-hungry AI data centers pose challenges to water-stressed cities

Some observers also point to the environmental costs associated with extremely power-hungry AI data centers.

The facilities generally house tens of thousands of servers operating simultaneously and advanced chips running at extremely high temperatures. The most advanced chips, the GPUs (graphics processing units) used for AI, can reach temperatures exceeding 90°C (194°F).

To operate optimally, therefore, data centers require vast quantities of stable, uninterrupted electricity as well as water for cooling their servers and other equipment.

Research fellow Joshi said that these demands pose a huge challenge for India's already water-stressed cities.

"These facilities need enormous amounts of water, constant electricity, and stable raw materials, and India has real problems on all three. What may attract companies more than the tax break is that India has few environmental regulations and is willing to hand over land quickly," said Joshi.

Jyoti Panday, Asia regional director of the Internet Governance Project, underscored the growing local opposition to data centers across the world.

"Globally, AI data centers are running into opposition over power use," she said.

"India's advantages, such as cheaper electricity, centralized decision-making and limited local resistance, make it attractive," she stressed, "but infrastructure constraints in cities like Mumbai and Chennai are real and could intensify."

Edited by: Srinivas Mazumdaru