Tata Communications plans to spend ₹4,350 crore on developing around 10,000 km of AI-ready connectivity infrastructure across India, as demand grows for high-speed, low-latency links between data centres and artificial intelligence hubs.
Announcing the investment on Monday, the company said the project would be rolled out in three phases between FY30 and FY36. The capital expenditure will be funded through operating cash flows.
The investment comes as India's operational data centre capacity reached nearly 1.8 GW in the first half of 2026, while an additional 3.86 GW is under construction or planned by 2030, according to commercial real estate services firm Cushman & Wakefield.
The expansion is aimed at capitalising on the growing need for data-centre-to-data-centre connectivity, driven by increasing cloud adoption, AI deployment and enterprise requirements for high-capacity digital infrastructure.
“From a business case perspective, it (monetisation) would be any industry benchmark 15 to 20% return on investment. As we bring capacity online, we will start to deliver revenues, so you start to kind of get that return,” Rupesh Chokshi, chief product and technology officer and executive vice-president & global business head, network services, Tata Communications, told Mint.
“We plan to fund it (the investment) from Tata Communications operating cash flows,” Chokshi said, adding that the company will have a first-mover advantage in data centre-to-data centre connectivity across the country as others are either focusing on creating more data centre capacities or on covering smaller locations in terms of connectivity.
First phase to cover Mumbai-Pune-Hyderabad-Chennai corridor
The initial phase of the project will focus on the Mumbai-Pune-Hyderabad-Chennai corridor, which Tata Communications considers an important growth route for AI infrastructure and hyperscale data centres.
Expected to become operational over the next three years, this phase will establish connectivity across the four cities. Later stages will extend the network to developing data-centre hubs in eastern and western India, including Vizag, Bhubaneswar, Jamnagar and GIFT City.
The company also plans to connect major data-centre clusters with international cable landing stations and integrate the new infrastructure with its global network backbone. This will facilitate connections between India's AI hubs and international markets.
Network to support hyperscalers and AI companies
Rupesh Chokshi, chief product and technology officer, executive vice-president and global business head for network services at Tata Communications, said AI workloads need reliable, resilient and low-latency connectivity, in addition to computing infrastructure.
The company is designing the network to cater to global hyperscalers, AI platforms, neocloud providers and enterprises. It will also continue upgrading its existing national long-distance and metropolitan networks to meet current connectivity needs among businesses and data centres.
Tata Communications operates a global network comprising more than 500,000 km of subsea optical fibre and over 200,000 km of terrestrial fibre.
The planned AI-ready network will leverage this existing infrastructure as the company looks to strengthen India's position in global cloud and AI connectivity.