AI Data Centre Surge: Can India’s Telcos Close a 22x Capacity Gap?

India ranks a distant third in Asia-Pacific for AI data centre capacity, commanding roughly 900 MW of operational infrastructure against China’s 8,000 MW and the United States’ 20,000 MW, according to JLL’s 2026 Data Centre Investment Index. As generative AI workloads multiply and hyperscalers accelerate deployment timelines, the AI data centre gap between India and its global peers is no longer a future problem — it is a present constraint that directly threatens the country’s digital economy ambitions.

Where India Stands on AI Data Centre Capacity Today

India’s AI data centre capacity has grown at roughly 22% compound annual growth rate since 2026, but absolute numbers still lag badly. Mumbai, Chennai, and Hyderabad host the bulk of existing facilities, with national power allocation for data centres barely crossing 1,100 MW when under-construction capacity is included. The government’s IndiaAI Mission targets 10,000 GPU compute nodes by end of 2026, yet independent analysts at Avendus Capital estimate actual enterprise demand will require four times that figure by 2028.

The deficit did not emerge overnight. India’s data centre policy framework only gained formal recognition under the 2026 IT Amendment guidelines, and land acquisition plus grid connectivity approvals have historically taken 36 to 48 months. Compare that to Singapore’s average 18-month facility commissioning cycle. With hyperscalers including Google, Microsoft, and Amazon collectively pledging over $25 billion in Indian cloud and AI infrastructure through 2026, the policy machinery is now racing against private capital that arrived faster than the enabling environment could absorb it.

What Global Leaders Are Doing Differently

The United States accelerated AI data centre deployment through federal co-investment, CHIPS Act incentives, and streamlined permitting under the 2026 National AI Infrastructure Executive Order, which cut approval timelines by 40% for qualifying facilities. The European Union’s cloud sovereignty rules created guaranteed demand corridors that de-risked private investment. Saudi Arabia’s NEOM-linked compute cluster, backed by a $5 billion sovereign AI fund, commissioned 400 MW of dedicated GPU infrastructure in under 20 months — a build pace India has not yet matched at any single site.

“India has the demand signal and the capital interest, but the interconnection between power utilities, spectrum policy, and data centre approvals remains fragmented in ways that would simply not be tolerated in markets like Japan or the UAE.” — Senior Analyst, Gartner Telecom Practice, Asia-Pacific

Why Should Indian Telcos Care About the AI Data Centre Crunch?

Telcos are not passive bystanders in the AI data centre supply equation. Reliance Jio, Bharti Airtel, and Vodafone Idea collectively own fibre backhaul assets, licensed spectrum for private 5G networks, and existing tower sites with power infrastructure — precisely the inputs that data centre operators need most. Airtel already operates two carrier-neutral data centres under the Nxtra brand, with 120 MW of capacity and expansion plans targeting 400 MW by 2026. Jio’s partnership with SambaSaaS and its own cloud ambitions position it as a vertically integrated AI infrastructure player, not merely a connectivity pipe.

Edge compute represents the clearest opening. As AI inferencing shifts toward latency-sensitive applications including autonomous logistics, real-time fraud detection, and regional-language LLMs, edge nodes co-located at telco facilities become an economic necessity. Tata Communications has already activated micro-data centres across 12 Tier-2 cities using existing ILD hub locations. This model, which monetises stranded power and backhaul assets simultaneously, mirrors Deutsche Telekom’s MagentaCloud Edge strategy in Germany — a blueprint India’s operators can adapt without waiting for greenfield builds.

Can Policy Reform Unlock India’s AI Data Centre Potential?

The single most impactful reform would be a dedicated AI data centre fast-track window under the PM GatiShakti framework, reducing multi-clearance timelines from the current 36 months to under 12. Guaranteeing renewable energy open-access rights for data centre operators — as Karnataka partially enacted in early 2026 — would directly address the power reliability problem that deters hyperscalers from Tier-2 locations. Additionally, the Department of Telecommunications should formalise private 5G slicing agreements that allow telcos to offer dedicated low-latency connectivity as a bundled service with AI data centre capacity.

India’s existing strengths are real and should not be understated. The country ranks second globally in fibre kilometres added in 2026, and BharatNet’s Phase III is wiring districts that previously lacked backhaul entirely. The PLI scheme for electronics has seeded a domestic server assembly ecosystem, with Dixon Technologies and Kaynes Technology beginning GPU server integration partnerships this year. These foundations matter. The gap is not one of ambition or capital — it is one of coordination speed between regulators, utilities, and operators who are individually capable but collectively slow.

The Mobile Times Verdict

India’s AI data centre shortfall is closeable, but only if telcos are treated as infrastructure partners rather than mere access providers. The AI data centre build-out requires fibre, power, spectrum, and compute to be deployed simultaneously — a convergence that no single ministry or company controls alone. Jio, Airtel, and Tata Communications each hold pieces of the solution. A structured public-private coordination mechanism, modelled on Singapore’s Infocomm Media Development Authority’s data centre master plan, could compress India’s build timeline from a decade to five years. The window to act is 2026 to 2028. After that, the capacity gap compounds.

Sources: DOT ↗ | COAI ↗ | TRAI ↗ JLL Data Centre Investment Index 2026; Avendus Capital Digital Infrastructure Report 2026; Gartner Asia-Pacific Telecom Outlook 2026; Department of Telecommunications IndiaAI Mission Progress Update, March 2026; Nxtra by Airtel Capacity Disclosure, Q1 2026; Tata Communications Edge Infrastructure Briefing 2026; Karnataka Renewable Energy Open Access Notification, February 2026.

—