On June 9, 2026, Meta announced an agreement with Reliance Industries to lease the first phase of an AI-enabled data center in Jamnagar, Gujarat: 168 megawatts, expandable later. This is Meta’s first AI data center in India, built to suit. Reliance provides end-to-end services — design, construction, renewable power, connectivity, and ongoing operations — while Meta covers the full cost of the energy and water its operations consume. The facility is expected to be ready within two years.
The core of this deal is not real estate; it is energy engineering. “We’re proud to be working with Reliance to build our first AI-enabled data center in India,” Mark Zuckerberg said. Reliance chairman Mukesh Ambani called the partnership “a transformative moment for India’s digital infrastructure.”
The Structure: Lease, Not Build
The most notable feature is the structure. Meta is not buying land and building; it is leasing capacity from Reliance. For Meta, capital expenditure converts into operating expenditure and delivery risk shifts to a local partner. For Reliance, it monetizes decades of power, water, and heavy-engineering capability from its refining and telecom empire against global AI demand. Meta’s release says the facility will support its global infrastructure and AI computing needs and complements the Project Waterworth subsea cable system — India is not just a market, it is a node in Meta’s global network.
The two companies are not strangers. Meta invested $5.7 billion in Jio Platforms in 2020, and in 2025 they launched a joint venture to bring Meta’s open-source AI models to Indian enterprises. This deal pushes the relationship from equity and licensing into physical infrastructure — and the Jamnagar campus is being developed by Reliance as one of the largest data center campuses in the world.
Clean Power, Seawater Cooling, and a Bill Meta Pays
The energy arrangement is the highlight. Alongside the data center, Meta announced nearly 1 gigawatt of new renewable contracts in India: 837 MW of solar and wind from CleanMax in Rajasthan and Karnataka, bringing their cumulative collaboration above 900 MW, plus 88 MW from Fourth Partner Energy across Tamil Nadu, Karnataka, Maharashtra, and Uttar Pradesh. The facility runs on renewable power, is cooled with desalinated seawater, and Meta pays the energy and water bills in full.
Putting water into the contract is not PR language. Water risk at data center sites is becoming a real cost line, and Jamnagar is Reliance’s refining heartland — desalination and power infrastructure already exist there. That is precisely why Reliance can offer a bundled package with power and water included.
Why India, Why Now
India is rapidly becoming a magnet for AI infrastructure. Microsoft has committed $17.5 billion by 2029; Amazon added $35 billion on top of plans totaling $75 billion; Google has a $15 billion AI infrastructure hub in the works; OpenAI secured 100 MW through Tata and is targeting a gigawatt. India’s data center capacity grew from roughly 375 MW in 2020 to about 1.5 GW in 2025, and KPMG projects more than 8 GW by 2030. Policy helps: foreign cloud providers serving overseas customers from Indian data centers enjoy tax exemptions through 2047.
After the announcement, Reliance shares jumped more than 2%.
Power Is the Hard Limit on AI Expansion
What does 168 MW mean in context? A large AI training campus runs from hundreds of megawatts into gigawatt territory, and the timeline for securing power has become the first bottleneck of AI expansion. Meta chose a site where power comes first: renewable contracts signed in lockstep with the data center, locking down energy risk before the concrete pours, while seawater cooling sidesteps freshwater disputes. Compared with labs scrambling to buy compute wherever it exists, what Meta is really buying here is certainty of energy supply.
Lessons for Infrastructure Strategy
Three observations. First, leasing rises: as AI demand outpaces self-build cycles, bundled “power and operations included” deals will multiply, and large-scale infrastructure capacity is being organized into a tradable service. Second, emerging markets compete on energy and land, not chips: India’s role is compute landlord, absorbing global training and inference workloads. Third, a note for product and platform teams: latency-sensitive services will still sit close to users, but the map for batch AI workloads is being redrawn — tax treatment, electricity prices, and water availability will shape your cost structure more than model choice.
Sources
- Meta Partners With Reliance on AI-Enabled Data Center in India — Meta Newsroom
- Meta signs first AI data center deal in India with Reliance — TechCrunch
- Meta partners Reliance for its first AI-enabled data centre in Gujarat’s Jamnagar — Moneycontrol
AI-assisted summary compiled from the sources above, reviewed by a human before publishing.