The TCS-OpenAI talks may signal India’s arrival into the physical spine of the global AI economy. Whether that arrival strengthens digital sovereignty or deepens strategic dependence will depend on the safeguards put in place as this new infrastructure takes shape.
For context, OpenAI, the company behind ChatGPT, is said to be negotiating to lease at least 500 megawatts (MW) of capacity from TCS’s HyperVault data-center project. The power would be used to train and operate its AI models locally, under what is being described as the arrival of “Stargate” in India, OpenAI’s blueprint for building massive computing infrastructure around the world.
For OpenAI, the commercial logic is straightforward. India already represents its largest user base outside the United States. Running models closer to where they are used means lower latency, better performance, and operational efficiency at scale. For TCS, the potential partnership offers something equally strategic: direct access to OpenAI’s leading models to build advanced AI services across core industries such as banking, retail, consumer goods, and manufacturing. If realized, it could accelerate TCS’s ambitions to position itself as a global leader in enterprise AI delivery.
This possible deal also fits squarely within a much larger investment wave sweeping across India. Google, Microsoft, and Amazon have each committed billions of dollars toward cloud and AI infrastructure in the country, drawn by India’s massive digital population and growing demand for computing capacity to train next-generation models. TCS, backed by TPG, is already developing a 1-gigawatt (GW) HyperVault facility, the very scale of infrastructure that hyperscalers and AI labs now require.
But the excitement around this build-out should not obscure its risks.
AI data centers consume enormous power. Facilities of this size draw hundreds of megawatts from the grid, a burden India’s energy system is not fully prepared to absorb. While renewable sourcing will be part of the mix, the sheer magnitude of demand means fossil fuels, including coal, will inevitably remain tied to the AI expansion, potentially complicating the country’s climate commitments.
Water use presents another growing pressure point. Large data centers rely heavily on water-based cooling systems, and water scarcity already affects many parts of India. Scaling AI infrastructure without addressing this constraint could deepen regional resource stress.
Finally, we must worry about control. When foreign technology companies build and operate the foundational layers of AI compute, ownership of the models, and control over the data pipelines running through them, largely remains offshore. India’s regulatory framework will need to ensure that data protection is enforced and that reliance on external models does not inadvertently limit the country’s own ability to develop independent, competitive AI ecosystems.
