Tata Communications plans to invest US$400–450 million, approximately Rs 4,000–4,350 crore, to build a 10,000-km fibre network connecting India’s major artificial intelligence (AI) and data centre hubs. The proposed network, called India’s AI Super Highway, will be developed in phases over the next three to five years to meet the growing demand for high-capacity, low-latency connectivity.
The first phase will focus on the Mumbai–Pune–Hyderabad–Chennai corridor, followed by expansion towards Visakhapatnam, Bhubaneswar, Bengaluru, Ahmedabad, Jamnagar and other data centre locations. The new infrastructure will connect with Tata Communications’ existing global network backbone and international cable landing stations.
The project is still in the planning and development stage. The company has announced the investment, but the proposed 10,000-km network is not yet operational.
Tata Communications AI Fibre Network: Investment and Timeline
Tata Communications announced the proposed investment on October 26, 2026. The company plans to fund the project through operating cash flows and develop the network in three phases, prioritising locations according to customer demand, traffic requirements and the growth of data centre infrastructure.
The investment is expected to support the construction of new long-haul and metropolitan fibre routes. The company has indicated that construction could begin in the fourth quarter of FY27, subject to vendor selection and other execution requirements.
| Project detail | Information |
|---|---|
| Project name | India’s AI Super Highway |
| Planned investment | US$400–450 million |
| Approximate investment in rupees | Rs 4,000–4,350 crore |
| Planned fibre network | Around 10,000 km |
| Development period | Three to five years |
| Funding source | Operating cash flows |
| Initial corridor | Mumbai–Pune–Hyderabad–Chennai |
| Expected meaningful revenue contribution | FY28–FY29 |
| Current status | Planned; not yet operational |
The investment will be deployed progressively rather than all at once. Revenue generation is also expected to increase as individual fibre routes are completed, activated and made available to customers.
Which Cities Will the AI Fibre Network Connect?
Tata Communications plans to expand the network in three phases to connect established technology centres and emerging data centre locations.
Phase 1: Mumbai, Pune, Hyderabad and Chennai
The initial phase will focus on the Mumbai–Pune–Hyderabad–Chennai corridor. These cities are important business and technology centres, with data centres, cloud infrastructure, enterprise customers and international connectivity routes.
The corridor is intended to provide high-capacity links between data centres and AI infrastructure. Such connections are increasingly important when computing workloads and data storage are distributed across multiple facilities.
Phase 2: Hyderabad, Visakhapatnam and Chennai
The second phase is planned to extend connectivity towards Visakhapatnam and strengthen links with Hyderabad and Chennai.
This expansion would extend the network into the eastern coastal data centre ecosystem and provide additional routes between technology hubs.
Phase 3: Eastern and Western India
The third phase is planned to expand connectivity across Bhubaneswar, Visakhapatnam, Bengaluru, Chennai, Mumbai, Ahmedabad and Jamnagar. The company also expects to serve emerging data centre ecosystems around Gujarat International Finance Tec-City (GIFT City) and other strategic locations.
The network is also intended to connect domestic data centre clusters with international cable landing stations, allowing customers to link their Indian infrastructure with overseas cloud and AI environments.
Why Does AI Need a Dedicated Fibre Network?
AI infrastructure depends on more than computing power. Data centres need reliable, high-capacity connections to exchange large volumes of data between servers, storage systems, cloud platforms and other computing facilities.
AI model training can involve moving large datasets between locations, while AI inference services must respond to user requests with minimal delay. When workloads are distributed across several data centres, network capacity, latency and route resilience can become important performance factors.
Tata Communications plans to design the new network around these requirements, rather than treating it simply as an extension of conventional enterprise connectivity.
The company says the infrastructure will be designed for high bandwidth, low latency and resilient routing. It has also highlighted a target round-trip latency of around 15–16 milliseconds for the Mumbai–Chennai route.
That figure is a planned network performance target, not a measurement from an operational network. Actual performance will depend on the final routes, equipment, network configuration and service conditions.
Who Will Use Tata Communications’ AI Super Highway?
The proposed network is aimed primarily at business and infrastructure customers rather than residential broadband users.
Potential customer groups include:
- Hyperscalers: Large cloud and technology companies operating extensive data centre infrastructure.
- Neocloud providers: Cloud infrastructure providers focused on specialised computing, including AI workloads.
- AI platforms and model developers: Organisations that need high-capacity links between computing and storage facilities.
- Software and technology companies: Businesses running data-intensive applications and distributed services.
- Global Capability Centres: Multinational companies operating technology and business operations in India.
- Large enterprises: Organisations expanding their use of cloud computing, AI and digital services.
Tata Communications has said it is discussing requirements with customers across these segments. However, the company has not announced final contracts for every proposed route, so the expected customer base should not be confused with confirmed bookings for the entire network.
The company has also indicated that the infrastructure will operate as a commercial, multi-tenant network. This means capacity is intended to be offered to multiple customers through commercial arrangements and longer-term contracts rather than being reserved exclusively for one company.
How Much Network Capacity Will the Project Add?
Tata Communications has described the proposed AI Super Highway as a major addition to its connectivity infrastructure. Its regulatory disclosures also outline plans for approximately 6 petabits per second of additional network capacity over FY2030 to FY2036.
The capacity addition is a longer-term target and should not be interpreted as capacity available to customers today. The new fibre routes will need to be built, equipped and activated before the planned services can be delivered.
The distinction is important because laying fibre does not automatically make a complete network operational. Capacity depends on the optical equipment deployed, the number of fibre pairs used, route engineering and the configuration of the wider network.
The proposed AI Super Highway will build on Tata Communications’ existing infrastructure, which includes more than 500,000 km of subsea optical fibre and over 200,000 km of terrestrial fibre, according to the company.
What Does This Mean for India’s Data Centre Industry?
The investment reflects a broader shift in telecom infrastructure demand. As cloud computing and AI workloads expand, data centre operators increasingly need dedicated connections between facilities, rather than relying only on traditional internet connectivity.
Tata Communications estimates that India’s data centre-to-data centre connectivity market could exceed US$1 billion over the medium term. This is the company’s market outlook, not a confirmed estimate of revenue that it will earn from the proposed project.
For India’s telecom and enterprise technology sectors, the project highlights the growing importance of long-distance fibre, metropolitan networks and international cable connectivity in supporting AI infrastructure.
The network could also create additional business opportunities for providers that can meet demanding customer requirements for bandwidth, uptime and latency. However, commercial success will depend on customer contracts, construction costs, execution timelines and the pace at which AI and data centre capacity expands.
When Will Tata Communications’ AI Fibre Network Become Operational?
Tata Communications plans to develop the network over three to five years, with meaningful financial contributions expected around FY28–FY29 as capacity is activated.
The company has indicated that construction could begin in the fourth quarter of FY27, subject to vendor selection. The final rollout will depend on execution, route planning, permissions and the pace of customer demand.
For now, the announcement represents a planned infrastructure investment rather than a completed network. Its significance will become clearer as construction begins, individual routes become operational and the company secures customer contracts.
The proposed 10,000-km AI Super Highway is designed to strengthen connectivity between India’s data centres, AI hubs and international networks. Whether it delivers the expected commercial returns will depend on how effectively Tata Communications converts the investment into operational capacity and long-term customer revenue.
