by SUSHANT KUMAR
If we are serious about “Digital India,” it must be our digital India, not a data colony wrapped in tricolour branding.
Across the world, the AI story is turning into one of anxiety – layoffs, restructuring, shrinking human roles. Every other week, Big Tech firms announce massive layoffs – Amazon is set to cut 30,000 jobs, owing mainly to efficiencies gained from AI, while Meta has cut around 600 roles from its superintelligence team. Back in June 2025, Amazon’s CEO Andy Jassy openly admitted that generative AI will “reduce” corporate headcount as automation scales.
But here in India, the same technology is being packaged as our next big leap forward. The irony is hard to miss. In a country still struggling for labour-intensive growth and equitable jobs, we’re celebrating automation as salvation.
The mirage of Google’s AI hub in Vizag
When Google announced its AI Hub in Visakhapatnam earlier this October, the press releases were drenched in optimism – “empowering India’s future,” “driving innovation,” “AI for Bharat.” Yet beneath that glow lies a more familiar script: a pattern of extraction and dependency that the Global South has seen for centuries.
Scholars Nick Couldry and Ulises Mejias call it data colonialism – the extraction of human life as digital raw material for profit. India’s new AI hubs risk becoming exactly that: vast data pipelines feeding Big Tech’s global models. The data of Indian citizens, institutions, and public systems will power algorithms refined and monetised elsewhere. The profits and patents will stay in the global North.
We’re told this is a partnership. But when the chips, servers, and cloud infrastructure are all imported, when the very code is written elsewhere, what kind of sovereignty is that? It feels uncomfortably close to the old colonial pattern: resources flowing outward, decision-making flowing upward.
Digital sovereignty or digital dependence?
The government’s rhetoric of “AI for Bharat” and “digital sovereignty” sounds bold, but the structural reality undercuts it. India still imports almost all high-end semiconductor hardware and relies on foreign cloud services for its data storage. Public sector digitization projects often run on private corporate clouds and opaque procurement contracts. Data governance frameworks remain weak, with limited public oversight.
So, while we speak of independence, we remain tethered to the architectures of foreign firms. What we’re building is not sovereignty, it is dependence dressed up in nationalist language.
The hidden environmental costs
AI hubs also come with physical costs rarely discussed in the hype. According to the International Energy Agency (IEA), global data centres consumed about 415 terawatt-hours (TWh) of electricity in 2024 – around 1.5% of total global demand – and could reach 945 TWh by 2030. The World Economic Forum estimates a 1 MW data centre can consume up to 25.5 million litres of water annually just for cooling, while global water use from data infrastructure could cross a trillion litres by 2030.
The Wire for more