Reliance unveils $110B plan for gigawatt AI data centers and nationwide edge compute in India

Mukesh Ambani has announced a ₹10 trillion (about $110 billion) push to build AI computing infrastructure across India over the next seven years—an industrial plan that treats “the cloud” as what it really is: land, power, cooling, and control.

According to TechCrunch, Reliance will build gigawatt-scale data centers, a nationwide edge-compute network, and AI services tied closely to its Jio telecom platform. Construction is already underway in Jamnagar, Gujarat, with more than 120 megawatts expected to come online in the second half of 2026. Ambani framed the project as technological self-reliance: India “cannot afford to rent intelligence,” he said, arguing that the true constraint on AI is scarce and expensive compute.

This is not a software story; it’s a utilities story with GPUs. Reliance says the build-out will be backed by 10 gigawatts of surplus green energy from solar projects in Gujarat and Andhra Pradesh. Hyperscale compute is a power-and-permits machine: grid connections, water rights, transmission upgrades, and local land politics. Whoever owns the data centers becomes a gatekeeper for the next layer of the economy, from manufacturing optimization to medical imaging to language models.

TechCrunch notes the investment lands amid a broader Indian capex race. Adani Group has outlined roughly $100 billion for AI data centers, and the Indian government expects more than $200 billion in AI infrastructure spending over the next two years. Global firms are also moving in: OpenAI is partnering with Tata Group on roughly 100 megawatts of AI capacity, with ambitions to scale to 1 gigawatt.

For India’s startups and universities, the promise is cheaper compute and “AI for Indian languages.” The risk is that compute becomes a quasi-public chokepoint controlled by a handful of conglomerates whose advantages are not algorithmic brilliance but regulatory endurance and access to capital. Ambani’s own history is instructive: Jio famously drove down mobile data prices, but it also concentrated market power and tethered vast swaths of Indian life to one corporate stack.

If India “cannot afford to rent intelligence,” it may instead end up renting it domestically—from national champions that look suspiciously like private ministries of compute. When AI capacity becomes strategic infrastructure, the next step is usually “national coordination,” which in most countries means subsidies, procurement favoritism, and rules that freeze incumbents in place.

Ambani is betting that whoever owns the megawatts owns the models. The rest of the economy will discover whether that is liberation from foreign cloud dependence—or simply a change of landlord.