Google Cloud plans more AI data centres to serve India demand

Google’s decision to expand AI‑ready data centres in India is a reminder that the AGI race is also an infrastructure race. Training and serving frontier‑scale models is ultimately a question of where power, land, connectivity and enterprise demand line up. By doubling down on India after committing $15 billion to a 1GW AI hub in Visakhapatnam, Google is betting that the next wave of AI usage—and the data that fuels it—will be geographically diversified rather than confined to US and European hyperscale regions.

Strategically, this move helps Google close the gap with Microsoft and AWS in a market where cloud adoption and AI workloads are ramping simultaneously. Gemini deployments for banks, hospitals, and consumer apps like Pocket FM don’t just drive revenue; they create embedded dependencies that will matter when more advanced multimodal and agentic systems arrive. For the broader ecosystem, Google’s “full stack” pitch—owning chips, models, cloud, and apps—competes directly with Microsoft’s similar narrative and leaves less oxygen for independent AI infra players in India. Over time, the density of AI‑heavy enterprises in markets like India could influence where cutting‑edge models are tuned, tested, and eventually co‑developed.