AirTrunk just made one of the biggest bets yet on India as the next home for AI compute. The Blackstone-backed, Australian data center operator said on Friday it will invest $30 billion in India by 2030, building out 5 gigawatts of new capacity, according to TechCrunch AI. That’s a massive commitment to a country whose entire data center footprint sits at roughly 1.5GW today.
What stands out here is the scale relative to where India is starting from. AirTrunk alone is promising more than three times the country’s current capacity.
What’s in the deal
- A planned 3GW data center at the Raigad Pen Growth Center in Maharashtra, tied to an investment of about ₹2 trillion (roughly $21 billion). The state just exchanged a letter of intent for the land.
- An existing pipeline of about 600MW across Mumbai, Chennai, and Hyderabad.
- A meeting between AirTrunk CEO Robin Khuda and Prime Minister Narendra Modi, who said the investment would strengthen India’s position as a global hub for cloud and AI.
The company wouldn’t confirm whether the Raigad project accounts for most of that 5GW target, or whether more sites are coming.
Why this matters
This is significant because it confirms a shift that’s been building all year: the AI infrastructure race is going global, and India is now a front-runner for the overflow. As research firm Bernstein notes via TechCrunch AI, India’s data center capacity could climb to as much as 8GW by 2030, up from about 1.5GW today.
New Delhi is actively courting that money. Earlier this year the government offered foreign cloud providers tax exemptions through 2047 on services sold overseas, as long as those workloads run from Indian data centers. That’s a long runway, and it’s clearly working.
AirTrunk isn’t alone. The list of companies pouring money into Indian infrastructure now includes Amazon, Google, Microsoft, OpenAI, and Uber, alongside domestic giants Reliance Industries, Adani Group, and TCS. When that many players move into the same geography at once, it tells you the compute crunch in established markets is real, and operators are hunting for land, power, and talent wherever they can find it.
The catch: power
Here’s where the optimism meets physics. Data centers are hungry for electricity, water, and land, and analysts keep pointing to power as the likely bottleneck. Deloitte estimates data center build-outs across the Asia Pacific region could require tens of terawatt-hours of additional electricity by the end of the decade.
Khuda says AirTrunk’s thesis rests on three things: government support, a deep pool of technical talent, and access to renewable energy. The renewable piece is the one to watch. Promising 5GW is one thing. Powering it cleanly and reliably in a country still expanding its grid is the harder problem, and it’s the same constraint slowing build-outs everywhere from Virginia to New York.
What to watch next
A few things will tell you whether this is momentum or just a headline number:
- Power agreements. Watch for renewable energy deals and grid commitments tied to the Raigad site. That’s the real proof of progress.
- Construction timelines. Land letters of intent are early steps. Shovels in the ground by 2027 would signal the 2030 target is credible.
- Competitive response. Expect Reliance, Adani, and the U.S. hyperscalers to keep escalating their own Indian commitments.
For practitioners and businesses watching where AI capacity is heading, the message is clear: India is moving from a back-office story to a core piece of global AI infrastructure. The capital is committing now. The power, talent, and timelines will decide who actually delivers.
More details are available in the original TechCrunch AI report.