Tata Consultancy Services CEO K Krithivasan has struck a notably bullish tone on the company’s push into AI data centres, revealing that TCS is in “advanced discussions” with multiple hyperscalers for new deals—building on its recently announced partnership with OpenAI.Speaking from TCS’s London offices, Krithivasan said India is staring at a massive supply gap when it comes to AI infrastructure. By his estimate, the country will need around 10 gigawatts of data centre capacity by 2030, but only 5–6 GW worth of projects have been announced so far.“There is going to be a lot of latent demand or unmet demand by 2030, so there is going to be a lot of investment required,” Krithivasan said.The OpenAI deal, announced at India’s AI Impact Summit in February, involves building facilities ranging from 100 MW to 1 GW. For a 1 GW data centre, TCS plans to invest about $1 billion toward infrastructure costs like racks, connectivity, power, and cooling—with partner TPG putting in an equal amount and the rest financed through debt. Total infrastructure costs alone could run between $7–8 billion.
From racks to AI agents, TCS wants to do it all
Krithivasan sees the data centre play as more than just a real estate bet. The end goal is to position TCS as a one-stop shop—offering everything from infrastructure and model training to AI agents and application intelligence. “The entire stack, we can offer to our customers,” he said.
TCS is also courting clients
At TCS’s London innovation hub—a sprawling space with views of the Thames—the company has been inviting current and prospective clients to brainstorm AI use cases. Visitors often start with Lovable, a popular vibe coding tool, to quickly prototype apps and test ideas on-site. One standout project built for TCS’s London Marathon sponsorship is a digital twin of a human heart, created using real athlete data, that can simulate how the organ responds to different training regimes.
