Can India deliver on AI ‘data city’ plan despite lingering water woes?

The project, centred on Visakhapatnam in Andhra Pradesh, is being promoted as a landmark project signalling India’s drive to expand its AI capabilities. 

State Information Technology Minister Nara Lokesh has said the development will anchor a technology ecosystem spanning about 100 kilometres and host up to 6 gigawatts of data centre capacity.

Major investors have signed on, with Google pledging around $15 billion for AI infrastructure in the state, while a joint venture between Reliance Industries, Brookfield and Digital Realty is investing a further $11 billion in a large data centre complex.

Lokesh has argued that Andhra Pradesh has planned for the heavy electricity and water demands of the sector, suggesting that excess monsoon water flowing into the Bay of Bengal could be channelled to cool facilities.

However, environmental experts have long warned that large data centres are among the most water-intensive pieces of digital infrastructure. Cooling high-density servers requires constant water circulation, much of which is lost to evaporation.

According to the Environmental and Energy Study Institute, a medium-sized data centre can consume up to 110 million gallons of water annually for cooling, while larger facilities may use as much as 5 million gallons a day — comparable to the needs of a small town.

That demand comes at a sensitive time for India, with UNICEF estimating that waterborne diseases cost the country roughly $600 million each year, with drought- and flood-prone regions affecting around a third of the population in recent years.

Two-thirds of India’s 718 districts face extreme water depletion, according to UNICEF. 

India is also the world’s largest user of groundwater, drawing supplies from more than 30 million access points. Groundwater provides 85% of drinking water in rural areas and nearly half of urban needs.

Beyond on-site consumption, data centres also carry indirect water costs. Power plants supplying electricity — particularly fossil fuel facilities — require substantial volumes of water for steam generation and cooling.

Manufacturing the semiconductor chips that power AI systems is also water-intensive, requiring ultrapure water in large quantities before equipment even reaches a data hall.

As India seeks to cement its place in the global AI race, the challenge will be balancing digital ambition with the realities of a country where access to safe and reliable water remains uneven.