Over 50% of Indian Firms to Add New Data Centre Capacity Within a Year as AI Workloads Surge

Mumbai — More than half of Indian enterprises plan to build new data centre capacity within the next 12 months, driven by the rapid growth of artificial intelligence (AI) workloads, according to a new report by Cisco released on Wednesday.

The study revealed that over 50% of Indian companies expect AI workloads to increase by more than 50% in the next three to five years. Meanwhile, 91% of Indian organisations have already deployed autonomous AI agents, though only 37% have the ability to secure them effectively.

Globally, 13% of organisations deploying AI are outperforming competitors by making key infrastructure investments that yield long-term advantages. Cisco calls these companies “pacesetters” — firms that have scaled AI deployment effectively to unlock use cases and deliver measurable returns on investment (ROI).

“These leaders build network-first foundations, prioritise power infrastructure, focus on continuous optimisation, and embed security from day one,” the report highlighted.

Simon Miceli, Managing Director for Cloud and AI Infrastructure across Asia Pacific, Japan, and Greater China at Cisco, warned that 45% of Indian companies risk developing what he termed ‘AI Infrastructure Debt’ — a form of technical debt that can lead to operational risks, security vulnerabilities, compliance issues, and competitive disadvantages if not addressed early.

The report identified four architectural principles that set AI leaders apart:

Continuous optimisation beyond deployment.

Network-first architecture as the foundation.

Early estimation of power needs.

Security built into infrastructure from the outset.

While most organisations concentrate on compute power, pacesetters prioritise network optimisation, with 81% rating their networks as “optimal” for AI workloads, Cisco noted.

“Pacesetters are winning not because they spend more, but because they made the right architectural bets early—before workloads demanded them, before bottlenecks appeared, and before security pressures intensified,” the report concluded.

With inputs from IANS