Tata Communications has placed a major bet on India’s artificial intelligence boom. On Monday, the Tata Group company said it will invest close to $450 million, or about Rs 4,350 crore, to build around 10,000 kilometres of AI-ready fibre across the country.
The company is calling the project India’s AI Super Highway. It is not a routine telecom upgrade. The network is being designed from the start for the heavy, constant traffic that moves between data centres, cloud platforms and AI systems, where delays matter and capacity needs keep climbing.
For anyone watching India’s data centre rush, the announcement answers a practical question: how will new facilities talk to each other, and to the rest of the world, at the speed AI work demands?
What Tata Communications announced
The planned network will run about 10,000 kilometres of new long-haul fibre and link India’s key AI and data centre corridors. Tata Communications said it will provide high-speed, low-latency connectivity between data centres, cloud platforms and enterprises, and connect those clusters to international cable landing stations and to its global network backbone.
The investment is put at $400 million to $450 million, and the company said it will fund the work through operating cash flows. The rollout is planned in three phases over three to five years.
In plain terms, this is long-distance plumbing for the AI economy. Long-haul fibre runs between cities and regions, and AI-ready means it is built for large, sustained data flows between machines.
Phase one starts with the busiest corridor
The first phase will focus on the Mumbai-Pune-Hyderabad-Chennai corridor, described by the company as one of India’s most important routes for AI and hyperscale data centre growth. Mumbai and Chennai are major landing points for undersea cables, Hyderabad is a fast-growing cloud and AI hub, and Pune sits in between with its own technology base.
Later phases will extend the network to Visakhapatnam, Bhubaneswar, Jamnagar and Gujarat’s GIFT City, along with other strategic locations on the east coast and in western India.
Who the network is built for
Tata Communications is aiming the network at the customers driving India’s data centre demand: global hyperscalers, neocloud providers, AI platforms, global capability centres and large enterprises. Banks and other organisations that move large volumes of data between their own facilities are also part of the intended customer base.
The network is being designed around direct data centre to data centre connections, with diverse routing and custom service levels. Diverse routing gives traffic more than one path. If one route has a problem, data can take another, which helps keep services running.
According to the company’s regulatory filing, the investment is expected to add around 6 petabits per second of capacity between FY2030 and FY2036, against current capacity of 1.25 petabits per second at 70 to 80 per cent utilisation.
Why this matters now
India’s data centre capacity is expanding quickly, but capacity alone does not make an AI hub. Training and running AI models involves moving huge amounts of data between processors, storage systems and data centres, often in real time. Slow links can become a bottleneck, even when computing power is available.
Ganesh Lakshminarayanan, managing director and chief executive of Tata Communications, said the company believes data centre to data centre connectivity in India could become a market exceeding $1 billion over the medium term.
“With our planned investments, we are building on our existing strengths to create one of the most comprehensive data centre connectivity platforms in the market,” he said. He added that the company sees a significant growth opportunity and wants to position itself to capture it.
For Tata Communications, the project also fits a longer shift in its business, from international voice and data services towards cloud, cybersecurity and enterprise digital services. A dedicated AI fibre network is hard to replicate quickly, because laying long-haul fibre takes time, permissions, equipment and capital.
What to watch next
Large fibre projects move in stages, and benefits usually arrive route by route. Investors and customers will watch construction milestones, which hyperscalers sign up, and whether data centre growth in newer hubs keeps pace.
If AI is going to run on Indian infrastructure, it will need Indian fibre to carry it. With this announcement, Tata Communications is betting $450 million that the demand is real, and that building the highway now is better than waiting for the traffic jam.
FAQs
How much is Tata Communications investing?
Close to $450 million, or about Rs 4,350 crore. The range given was $400 million to $450 million, funded through operating cash flows.
How long will the network be?
Around 10,000 kilometres of new long-haul fibre across India, built in three phases over three to five years.
Which cities will be connected first?
The first phase focuses on the Mumbai-Pune-Hyderabad-Chennai corridor. Later phases will extend to Visakhapatnam, Bhubaneswar, Jamnagar, GIFT City and other strategic locations.
What is the AI Super Highway?
It is the name Tata Communications has given the planned network: a high-speed, low-latency fibre fabric linking India’s key AI data centre corridors, cable landing stations and its global network backbone.
Who will use the network?
Hyperscalers, neocloud providers, AI platforms, global capability centres, enterprises and other large organisations that need high-capacity links between data centres.
That is the story so far. Informeia will follow the rollout as the first corridor takes shape.
