Adani Group is preparing to thread artificial intelligence through its sprawling portfolio, with the group outlining a broader strategy that combines AI infrastructure, data centers, energy and digital capabilities. The Adani AI strategy rests on a simple premise: whoever controls dependable compute capacity will shape how Indian industry adopts the technology. Three major data center developments in Navi Mumbai, Hyderabad and Vizag form part of AdaniConneX’s expanding infrastructure footprint, with the current portfolio listing potential capacities of 1 GW in Navi Mumbai, 600 MW in Hyderabad and 1 GW in Visakhapatnam. Partnerships with hyperscalers including Google and Microsoft give parts of the buildout established commercial relationships, while AdaniConneX continues to expand capacity across its data center portfolio. Rakesh Dhanda, chief digital officer at Adani Group, laid out the thinking to ETDatacenters, and his remarks read more like a statement of industrial positioning than a product pitch.

Adani AI strategy turns demand into infrastructure

Dhanda framed the moment in terms of demand, and the logic behind that framing explains the urgency across the group. Dhanda told ETDatacenters, “Data centers are growing exponentially considering AI, which will be the future across businesses.” That reading reflects the infrastructure demands created by AI workloads, which have pushed data center operators and technology companies to plan larger amounts of power and compute capacity. The group’s answer is to build data center capacity at a scale that places compute infrastructure at the center of its broader digital and infrastructure strategy. Adani already builds heavy infrastructure at scale, and a data center campus fits that operating instinct. Investors should watch whether that instinct converts into delivery speed, which the sections below test.

Three gigawatt campuses anchor the compute bet

The group’s current project disclosures show how large it intends to build, with AdaniConneX listing potential capacities of 1 GW in Navi Mumbai, 600 MW in Hyderabad and 1 GW in Visakhapatnam. AdaniConneX’s current portfolio includes data center developments in Navi Mumbai, Hyderabad and Visakhapatnam, alongside facilities in Chennai, Noida and Pune.AdaniConneX currently lists potential capacity of 1 GW in Navi Mumbai, 600 MW in Hyderabad and 1 GW in Visakhapatnam, while Adani has separately announced partnerships involving Google and Microsoft. Gigawatt-scale capacity signals an ambition of a different order from individual megawatt-scale facilities, particularly as AdaniConneX develops large-scale projects in Navi Mumbai and Visakhapatnam. Developments across Navi Mumbai, Hyderabad and Visakhapatnam would give AdaniConneX a geographically distributed data center footprint across major Indian markets. Power availability, construction progress and the commissioning of individual phases will determine how quickly planned capacity becomes operational capacity.

AdaniConneX has developed data center facilities in Noida and Pune, with Phase 1 at Noida listed as live from January 2025 and Phase 1 at Pune listed as live from October 2025. That detail shows the group has moved beyond planned projects into operating and customer-handover stages, with Adani Enterprises reporting more than 55 MW of operational data center capacity in April 2026. Dhanda tied the buildout to national policy and to the wider AI boom. He told ETDatacenters, “You require data centers to be set up in India and the government is also promoting the data center business here, considering the AI boom.” Government support and state-level approvals have helped advance data center investments, although individual projects still require separate land, power, construction and regulatory clearances.

Why hyperscaler tenants change the risk math

Documented relationships with Google and Microsoft strengthen the commercial context around parts of Adani’s AI infrastructure buildout, although those relationships should not be treated as confirmed customer contracts for every AdaniConneX project. Long-term customer commitments can support financing for capital-intensive data center developments, although the contractual terms of individual AdaniConneX customer arrangements have not been publicly detailed in the sources reviewed. Hyperscalers are expanding AI and data center capacity in India as demand for cloud and AI services grows, while India’s regulatory framework also creates specific requirements around the handling and storage of certain categories of data. A developer that combines data center development, infrastructure delivery and connectivity can offer hyperscalers an integrated route to deploying capacity, although deployment speed still depends on project-specific approvals and construction milestones.

AdaniConneX brings a global operator to the table

Adani Group runs its data center business primarily through AdaniConneX, a 50:50 joint venture between Adani Enterprises Limited and the US-based EdgeConneX. EdgeConneX operates nearly 90 data centers across 20 countries, which hands the Ahmedabad-based group an operating playbook it did not have to invent from scratch. That pairing combines Adani’s infrastructure capabilities with EdgeConneX’s global data center design, engineering and operating experience. As a joint venture, AdaniConneX operates through a partnership structure in which Adani Enterprises and EdgeConneX each hold a 50% stake. Nothing in Dhanda’s comments points to friction on either count. Investors will want clarity on how capital flows between the partners as three gigawatt campuses move from plan to construction.

Adani AI strategy reaches into the EPC value chain

Dhanda made clear that the Adani AI strategy extends well beyond selling capacity to others. He said the group uses AI for digital transformation and embeds the technology in its business strategy. On project delivery, Dhanda told ETDatacenters, “If you talk about project management, we are leveraging AI starting from the planning phase and pre-construction phase — a complete EPC value chain; then we have O&M operations and handover management.” Engineering, procurement and construction is where large infrastructure programs gain or lose months, which means applying AI from the planning stage targets the point where value leaks most. Operations, maintenance and handover close the loop, and data gathered in those phases can feed the design of the next project. A group that builds gigawatt campuses and uses AI to build them becomes both a supplier and a customer of the technology.

The group operates across a broad portfolio of infrastructure, energy and other businesses, and Adani says it is deploying AI across multiple business processes and operating activities. The group has described AI as more than a standalone technology intervention, linking it to changes in how infrastructure and operating activities are delivered. Dhanda said, “It (AI) is not only about technology intervention, but it is also a new way of delivering projects and solutions to customers, as well as for internal capital purposes.” Adani Group previously partnered with Autodesk on a three-year strategic program focused on improving the lifecycle of its infrastructure projects through digital tools and advisory services. The two companies signed a three-year contract to deploy advanced digital tools and advisory services, aiming to lift efficiency and innovation across the infrastructure portfolio.

Odisha approval widens the Adani data centers footprint

The group’s ambitions in the sector extend beyond the three named campuses. High Density Data Centre Ltd, a subsidiary of Adani Infra (India) Ltd and part of the Adani Group, recently received approval from the Odisha government to develop an integrated data center park. The investment for that park comes to about ₹1.04 lakh crore. Meanwhile, the state government approved three other data center and AI-related projects with a combined investment of ₹38,391 crore. Those four projects bring total investment across Odisha to nearly ₹1.42 lakh crore. Numbers of that size mark the state as an emerging destination for compute infrastructure.

An integrated data center park differs from a single building: it bundles land, power, connectivity and multiple halls into one planned site. Such parks allow phased delivery, and capacity arrives in tranches as customers commit. The Odisha approval gives the group a pipeline that stretches beyond the current three-campus program and into a fourth region. Approvals do not equal energized racks, and schedules will depend on grid access, water and equipment supply. The group will need to secure those inputs across several states at the same time. Success there would turn a portfolio of announcements into a durable operating footprint.

Viksit Bharat and the capital spending backdrop

Dhanda tied the group’s plans to national ambition. He told ETDatacenters, “If you look at the Viksit Bharat (Developed India) vision, which is to develop the country by 2047, India is investing in capital infrastructure.” “This year, India is investing $130 billion in CapEx, and the Adani Group is an integral part of India’s growth strategy,” the top executive added. Public capital spending of that magnitude creates demand for digital infrastructure alongside roads, ports and power; every modern asset generates data that needs a home. Adani Group positions itself as a builder for that spending cycle rather than a bystander. The 2047 horizon signals that the group thinks in decades, which suits assets with long payback periods.

What the Adani AI strategy means for the compute race

India’s compute race will reward operators that can deliver power, land, capital and customers at once. Adani Group holds credible claims on several of those pieces, from hyperscaler relationships to an experienced joint venture partner and state approvals worth lakhs of crores. Therefore the Adani AI strategy deserves close attention from infrastructure investors, cloud buyers and policy makers alike. Execution risk remains real: gigawatt campuses strain grids, supply chains and construction crews at the same time. Watch the operating milestones for Noida and Pune, the December 2026 RFS target for Navi Mumbai, the existing operational capacity in Hyderabad and the December 2028 Phase 1 RFS listed for Visakhapatnam. Finally, the group’s use of AI and digital tools across its infrastructure and operating activities adds another dimension to its broader strategy of building and operating AI infrastructure.