Tata Communications to Invest Up to $450 Million in 10,000 Km AI Fibre Network Across India

Tata Communications said on Monday that it will invest up to USD 450 million to build a high-speed, low-latency fibre network across India, aimed at linking the country's data center clusters and artificial intelligence hubs. The company described the project as an AI superhighway and said it will lay around 10,000 kilometers of new long-haul fibre, one of the largest AI-ready connectivity builds in the country.

The announcement puts fiber optics technology at the center of India's push to support AI workloads. As cloud adoption accelerates and enterprises deploy AI applications, demand is rising for local data storage and processing, and for the ability to move large volumes of data quickly between facilities.

Where the network will run

The first phase will focus on the Mumbai-Pune-Hyderabad-Chennai corridor, a major route for AI and hyperscale growth. Later phases will extend into emerging data center ecosystems on the east coast and in western India, including Visakhapatnam, Bhubaneswar, Jamnagar and GIFT City, the company said. The network will also connect to Tata Communications' existing global backbone and to international cable landing stations, linking domestic infrastructure with overseas AI hubs.

The company operates more than 500,000 kilometers of subsea optical fibre and over 200,000 kilometers of terrestrial fibre. It plans to add around 6 petabits per second of network capacity under the program, taking total capacity to approximately 7.25 pbps once the expansion is completed. Existing network capacity utilization stands at around 70% to 80%. The capacity addition is scheduled between FY30 and FY36 and will be financed through operating cash flows.

The customers in line

Tata Communications said the network is designed for hyperscalers, AI platforms, cloud providers, global capability centers, enterprises and digital-native businesses. Managing Director and Chief Executive Officer Ganesh Lakshminarayanan said the India data center-to-data center connectivity market could exceed USD 1 billion over the medium term. He added that an increasing share of the company's expansion is being driven by anchor-customer commitments and contracted demand. The company did not give a completion timeline for the project.

Rupesh Chokshi, Executive Vice-President and Global Business Head for Network Services, said the AI economy needs more than compute and requires a network built for speed, scale and resilience. The company reported consolidated net profit attributable to equity holders of Rs 341.03 crore for the second quarter of the current financial year. Its shares closed the session 1.56% higher at Rs 1,680.

What it means for technical careers

For engineers and technicians, the build points to sustained demand for skills in optical transport, long-haul route planning, data center interconnection and network automation. The phased schedule stretching into the mid-2030s suggests that work will be spread across planning, construction and operations teams rather than concentrated in a single hiring cycle. As AI workloads move between training sites, inference locations and end users, specialists in latency measurement and capacity management are likely to be drawn into projects of this kind.

The investment lands in a busy stretch for India's digital infrastructure. E2E Networks separately plans to add 5.4 MW of AI compute capacity, according to the same reporting, while access to computing capacity and reliable power remain central concerns for AI projects. Analysts have noted that the ability to move data quickly between facilities is becoming as important as the chips themselves.

What to watch next is how much of the planned capacity is backed by signed customer deals before construction begins, and whether later phases reach the emerging clusters on the east coast on the same schedule. Also worth tracking is whether rival operators respond with their own long-haul builds, and how quickly the corridor moves from announcement to usable commercial service.