Gujarat Data Centre Policy 2026-29 Explained: ₹6 Lakh Crore Push for 7.5 GW Green AI Hub

The basic concept: A data centre is a dedicated physical facility housing thousands of networked servers, storage systems and networking equipment that store, process and distribute digital data. Every email, UPI payment, video stream, government e-service and AI chatbot query is ultimately processed inside a data centre somewhere.

Key components: Beyond servers, a data centre requires massive uninterrupted electricity supply, sophisticated cooling systems (since servers generate enormous heat), high-speed fibre connectivity, and physical and cyber security. Reliability is measured in "uptime", with top-tier facilities designed for near-continuous availability.

Hyperscale and AI data centres: A hyperscale data centre is an extremely large facility — typically run by or for global cloud and technology majors — designed to scale to hundreds of megawatts of IT capacity. AI-ready data centres are a newer breed built around dense clusters of specialised chips (GPUs) for training and running artificial intelligence models; they consume far more power per rack and need advanced cooling, which is why capacity is now measured in gigawatts of power rather than just floor area.

Why states are competing for them: Data centres bring large capital investments, anchor cloud and AI ecosystems, improve digital service quality by keeping data closer to users (lower latency), and support data-residency needs of sectors like banking. They are, however, not large direct employers — their value lies in the digital ecosystem they enable.

Targets and vision: The policy aims to make Gujarat India's preferred destination for data centres by enabling up to 7.5 GW of capacity and attracting about ₹6 lakh crore of investment in its first phase. The state projects itself as a hub for cloud computing, AI, e-commerce, digital governance and smart manufacturing, and treats data centres as strategic infrastructure.

Gujarat's claimed edge: The Secretary of Gujarat's Department of Science and Technology told The Indian Express that Gujarat is the first state to announce the largest hyperscale capacity and the first with a dedicated policy specifically for green AI data centres, arguing that other states handle projects case-by-case. Gujarat's underlying strengths include about 69 GW of installed power capacity, of which 47 GW is renewable, plus ready industrial ecosystems at Dholera and GIFT City.

Early response: According to state ministers, proposals for nearly 10 GW have already been received from around 14 hyperscale investors, including companies partnering global technology firms — evidence, the state argues, of demand exceeding the policy's own target.

Fiscal incentives: The package includes exemption from stamp duty and registration fees for leasing or buying land; 100% reimbursement of electricity duty for 20 years; a power subsidy of ₹1 per unit consumed for 20 years from commencement of commercial operations; reimbursement of eligible expenses on plant and machinery, building infrastructure and services for up to 20 years; and, per official releases, capital and interest subsidies, power tariff support and SGST reimbursement.

The Dholera sweetener: To concentrate the industry in the Dholera region, a capital subsidy of 2.5% on eligible fixed capital investment is offered — covering core technical and infrastructure expenses but explicitly excluding the cost of land, land development and chips (servers/GPUs), which typically form the biggest cost head.

Non-fiscal facilitation: Relief in Floor Space Index, relaxed building norms, assured uninterrupted power and water, nodal officers for single-window facilitation, and faster approvals round out the ecosystem measures.

The power problem: AI-era data centres are extraordinarily power-hungry — a single hyperscale campus can consume as much electricity as a mid-sized city. Concentrating gigawatts of new demand on a fossil-heavy grid would strain supply and raise emissions. The policy therefore mandates that at least 51% of electricity for core operations come from green and renewable sources, leveraging Gujarat's solar, wind and hybrid parks, including the giant Khavda renewable energy park in Kutch.

The water problem: Data centres need large volumes of water for cooling. Gujarat is a water-stressed state, so the policy requires and financially supports captive desalination plants, ensuring the sector does not divert freshwater meant for agriculture, industry or drinking. The Reliance-Meta Jamnagar facility, cooled with desalinated seawater, previews this model.

Industry alignment: The green mandate matches where global capital is headed — Google's planned AI campus in Andhra Pradesh is to run entirely on clean energy, and hyperscalers such as Meta contract renewable power (nearly 1 GW of new Indian clean energy in Meta's case) to meet corporate net-zero commitments.

India's first greenfield smart city: Dholera Special Investment Region (SIR), located about 100 km southwest of Ahmedabad along the Delhi-Mumbai Industrial Corridor, is India's first greenfield smart industrial city, with plug-and-play infrastructure, large contiguous land parcels and planned utilities.

Emerging tech cluster: Dholera already hosts India's first commercial silicon semiconductor fab (Tata Electronics-PSMC, scheduled for commissioning in 2028), creating a natural synergy between chip-making and data infrastructure. The state envisions Dholera as potentially the world's largest "data centre city", supported by a new airport and a proposed semi-high-speed rail link to Ahmedabad, and plans to develop it as a hub for Global Capability Centres (GCCs) as well.

The deal: In June 2026, Reliance Industries and Meta announced an agreement under which Reliance will build a 168 MW AI-enabled data centre at Jamnagar, Gujarat — to be delivered within about two years — which Meta will lease with options to scale. It is Meta's first AI-enabled data centre in India and India's first built-to-suit facility for a global hyperscaler.

Green template: The facility will run on renewable energy and be cooled with desalinated seawater, with Meta covering the full cost of energy and water. Meta has separately contracted nearly 1 GW of new Indian clean energy through CleanMax (837 MW) and Fourth Partner Energy (88 MW). The project extends a partnership dating to Meta's $5.7 billion investment in Jio Platforms in 2020 and a 2025 enterprise AI joint venture.

The demand-capacity mismatch: India generates nearly 20% of the world's data but hosts only around 3% of global data centre capacity — a gap that represents both a vulnerability and an enormous investment opportunity as AI adoption, 5G, digital payments and data-residency requirements expand.

Policy evolution: Telangana pioneered a state data centre policy in 2016, with Tamil Nadu, Uttar Pradesh, West Bengal, Odisha, Maharashtra, Haryana and Andhra Pradesh following. At the Union level, data centres were included in the harmonised master list of infrastructure sub-sectors in the 2022-23 Budget, easing access to long-term credit, and MeitY has worked on a national data centre policy framework.

Drivers of localisation: The RBI's 2018 directive requiring payment system data to be stored in India, the Digital Personal Data Protection Act, 2023, sectoral data-residency norms, and the IndiaAI Mission's push for domestic AI compute capacity all reinforce demand for locating data infrastructure within India. Mumbai and Chennai (cable-landing hubs), the NCR and Hyderabad have been the traditional clusters; Gujarat is now bidding to leapfrog them on the strength of cheap renewable power, land and coastal desalination.

Grid and transmission stress: Adding several gigawatts of round-the-clock demand requires matching investment in renewable generation, storage and transmission; the 51% green mandate is only as good as the firm (24x7) green power available.

Water and environment: Desalination is energy-intensive and produces brine discharge, requiring careful coastal environmental management.

Import dependence: The most expensive components — GPUs and servers — are imported and explicitly excluded from subsidy calculations, meaning much of the capital expenditure value flows abroad until domestic electronics and semiconductor manufacturing matures.

Limited direct employment: Data centres create relatively few permanent jobs relative to investment size; the real employment dividend depends on the wider AI, cloud and GCC ecosystem they attract.

Inter-state competition: With many states offering competing subsidies, there is a risk of a "race to the bottom" in incentives without commensurate ecosystem returns.

The policy's success will hinge on execution rather than announcement. Gujarat must ensure firm round-the-clock green power through storage-backed renewable contracts so the 51% mandate is met in practice, fast-track transmission corridors to Dholera and coastal sites, and enforce environmentally sound desalination with brine management norms. Converting the ~10 GW proposal pipeline into commissioned capacity will need genuinely time-bound single-window clearances and reliable last-mile infrastructure. For the country as a whole, state incentives should be complemented by domestic manufacturing of servers, networking gear and, eventually, AI chips — linking the data centre boom with the Semicon 2.0 and electronics component programmes — so that India captures the full value chain, not just the buildings and electricity bills, of the global AI infrastructure race.