India is facing a critical electricity shortage as the rapid expansion of AI-driven data centers outpaces the national power grid's capacity.

This energy crisis threatens to become the primary bottleneck for India's artificial intelligence ambitions. Because AI workloads require high-performance computing chips that consume massive amounts of power, the infrastructure needs of these facilities are beginning to clash with the stability of the existing electrical grid.

Current data center power capacity in India stands at 1.5 GW [1]. However, that figure is projected to surge to between eight GW and 10 GW by 2030 [1]. To put this scale into perspective, the electricity consumption of a single large data center is often comparable to that of a small city [1].

Private investors and the government are attempting to bridge this gap through massive capital injections. The Adani Group has planned a total spend of $100 billion [2] to support the nation's AI goals. As part of this expansion, the group has sought up to $5 billion [3] for a Google data center in India.

These local efforts mirror a global trend in infrastructure spending. Forecasts suggest that global AI-driven data center capital expenditure will reach $5.2 trillion by 2030 [4].

Despite the available capital, the physical reality of power distribution remains a hurdle. The high-performance chips required for AI processing demand constant, high-voltage power that the current grid is not fully equipped to handle. This has forced operators and power utilities to rethink how energy is delivered to these new campuses.

“The electricity consumption of a single large data center is often comparable to that of a small city.”

The transition from traditional cloud computing to AI-heavy workloads represents a fundamental shift in energy requirements. For India, this means that AI leadership cannot be achieved through software or investment alone; it requires a comprehensive overhaul of the national energy grid to prevent industrial growth from causing widespread power instability.