Two Conglomerates, One Trillion-Rupee Premise
At the India AI Impact Summit in New Delhi — the country's first turn hosting a global AI summit, running February 16–21, 2026 — Reliance and Adani disclosed AI infrastructure commitments within 48 hours of each other. Adani went first, on February 17, pledging $100 billion through 2035 to build renewable-powered, AI-ready data centres. Mukesh Ambani followed on February 19, committing Reliance and Jio to invest roughly $110 billion (₹10 trillion) over seven years. Between them, the two groups put more new private capital behind India's AI build-out in a single week than most G20 nations spend on digital infrastructure in a decade.
Both pledges rest on the same argument: that India cannot depend indefinitely on renting compute from the same handful of American hyperscalers that already serve the rest of the world, and that the country's edge lies in pairing abundant, cheap renewable power with a market of over a billion eventual AI users. Where the two groups diverge is in how they propose to capture that opportunity — and that divergence is the more useful story for anyone trying to read where Indian capital is actually flowing.
RIL's most recent disclosure of the build-out came at its 49th Annual General Meeting on June 19, 2026, where Chairman Mukesh Ambani folded the AI plan into the company's broader FY26 results and the long-awaited filing of Jio Platforms' draft IPO prospectus with SEBI on the same day — a sequencing that was almost certainly deliberate.
Reliance's Playbook: Vertical Integration, Telecom to Token
Reliance's approach to AI is an extension of the same playbook that built Jio: own the full stack, from spectrum to silicon to subscriber, and subsidise scale with patient capital from the group's energy and retail cash flows. The vehicle is Reliance Intelligence Limited, incorporated as a wholly-owned subsidiary on September 9, 2025, after Ambani first outlined it at the 48th AGM. Its stated mandate spans four roles: building gigawatt-scale AI data centres, supplying India's edge-compute backbone, delivering consumer- and enterprise-facing AI services, and housing the research and engineering talent to do all three.
The flagship asset is a multi-gigawatt, green-energy-powered AI campus under construction in Jamnagar, Gujarat, with an initial 120-megawatt phase targeted to come online in the second half of 2026. Reliance has paired the build-out with hyperscaler partnerships rather than going it alone: a deepened AI alliance with Google announced in August 2025 that bundles Gemini access for Jio subscribers, and a newly signed joint venture with Meta to co-develop an AI-enabled data centre at the same Jamnagar site, announced June 10, 2026 — just nine days before the AGM.
On the consumer side, the 49th AGM introduced two concrete products rather than another roadmap slide: Jio Teleframe, a platform for deploying AI agents, and Jio Call Agent, a native AI voice assistant. Akash Ambani, Chairman of Reliance Jio, framed the underlying logic plainly in RIL's own Q4 FY26 results commentary: Jio's connectivity and edge-compute infrastructure positions it as the channel through which AI reaches Indian consumers, much as it once did for mobile data.
What makes this credible rather than aspirational is the balance sheet behind it. RIL closed FY26 (year ended March 31, 2026) with record consolidated revenue of ₹11.76 lakh crore ($124.0 billion), up 9.8% year-on-year, and annual capital expenditure of ₹1.44 lakh crore ($15.2 billion) — already one of the highest capex runs of any Indian corporate, now being partly redirected toward AI. Net debt to EBITDA stood at a conservative 0.60x, leaving meaningful headroom to fund the AI build without straining the balance sheet that funds O2C, retail and the rest of the group.
Adani's Playbook: Energy as the Moat
Adani is not trying to build a consumer AI brand. Its bet is structural and, in its own words, "energy-compute symmetry" — the idea that the binding constraint on AI growth globally is not chips or talent but reliable, affordable power, and that whoever controls gigawatt-scale renewable generation controls the AI supply chain's chokepoint. Gautam Adani put it bluntly in the group's own February 17 release: nations that master the balance between energy and compute will shape the next decade, and India intends to be a creator of intelligence rather than merely a consumer of it.
"India will not be a mere consumer in the AI age. We will be the creators, the builders and the exporters of intelligence."Gautam Adani, Chairman, Adani Group — Official Media Release, adani.com, 17 Feb 2026
The mechanism is AdaniConneX, the group's existing data-centre platform built in partnership with EdgeConnEx, which Adani is scaling from roughly 2 gigawatts of national capacity today toward a 5-gigawatt target by 2035. The $100 billion direct investment is explicitly designed to catalyse a further $150 billion from manufacturing, sovereign cloud and supporting industries, which the group projects will build a $250 billion AI infrastructure ecosystem in India over the decade — a multiplier effect Reliance's plan does not claim in the same terms.
Crucially, Adani is positioning itself as infrastructure landlord to the industry rather than a single integrated operator. Partnerships announced alongside the $100 billion plan include Google's largest gigawatt-scale AI data-centre campus in India, sited at Visakhapatnam; Microsoft-backed campuses in Hyderabad and Pune; and an expanded joint venture with Flipkart for a second high-performance AI and e-commerce compute facility. The entire platform leans on Adani Green Energy's 30-gigawatt Khavda renewable project in Gujarat — already India's largest single renewable site, with more than 10 gigawatts operational — alongside a further $55 billion earmarked for renewable capacity and battery storage, including what the group describes as one of the world's largest single-location battery energy storage systems.
Capital, Capacity and the Execution Gap
Strip away the headline numbers and the two strategies imply very different execution risk. Adani's AdaniConneX platform already has roughly 2 of its 5-gigawatt target operating today — a genuine head start in physical capacity, even if most of that capacity predates the AI-specific framing. Reliance's flagship Jamnagar campus, by contrast, is earlier-stage: its first disclosed phase is a comparatively modest 120 megawatts, with the gigawatt-scale ambition still under construction as of mid-2026.
The financing models also diverge in ways that matter for risk. Reliance's $110 billion is funded substantially from a single, diversified, cash-generative balance sheet — the same one that produced record FY26 EBITDA of ₹2.08 lakh crore — giving Ambani more direct control but also concentrating execution risk inside one company. The pending Jio Platforms IPO, whose draft prospectus was filed with SEBI on the same day as the 49th AGM, is widely read as a parallel mechanism to crystallise value from the digital business and free up further capital for the AI build, rather than a sign the core balance sheet is stretched.
Adani's $100 billion, by contrast, is explicitly designed to catalyse roughly 1.5 times that amount in co-investment from manufacturing and cloud partners — a model that requires less direct capital from Adani itself but depends more heavily on the willingness of third parties (Google, Microsoft, Flipkart, and unnamed others still in discussion) to commit alongside it. Both models carry a common, longer-dated risk that is easy to understate amid the announcements: a seven-to-nine-year build-out horizon assumes AI compute demand keeps compounding at something close to today's rates, and that GPU and grid-connection supply chains do not become the new bottleneck in their place.
What This Means for India's AI Sovereignty
Set against the wider picture, $210 billion of combined private commitment is large but not dominant. It compares with an estimated $200 billion-plus the Indian government expects in total AI infrastructure spending over the next two years, and with the more than $630 billion in capital expenditure U.S. technology giants are expected to deploy globally in 2026 alone, by outside estimates cited at the summit. India's structural argument — land, sunshine, and a captive renewable cost advantage that the West largely lacks — is real, but it is a cost edge, not yet a technology edge; both Reliance and Adani remain dependent on Google, Microsoft, Meta and Nvidia-class silicon for the layers above raw power and floor space.
The public layer matters here too. Alongside the corporate pledges, the government's own IndiaAI Mission outlined plans to add roughly 20,000 GPUs to its subsidised national compute pool — a complementary, much smaller-scale public option that startups and research institutions can draw on without needing access to either conglomerate's infrastructure. Both Reliance and Adani have publicly committed to reserving a share of their compute for domestic AI startups and academic researchers, which, if honoured, would meaningfully widen who actually benefits from this build-out beyond the two groups' own ecosystems.
For India specifically, the more interesting outcome may not be which conglomerate "wins," but whether the two strategies end up complementary rather than competitive: Adani supplying gigawatt-scale, renewable-powered compute as a wholesale utility to hyperscalers and enterprises, while Reliance verticalises that compute into consumer and small-business AI products distributed over Jio's network. That division of labour — energy infrastructure versus retail AI distribution — would mirror how the two groups already coexist in adjacent sectors like ports-and-logistics versus telecom-and-retail, rather than forcing a head-to-head contest neither may need to win outright.
The Investor Lens: Two Different Risk Profiles
For Reliance, AI is incremental capex layered onto an already-diversified cash machine; the bigger near-term swing factor for the stock is probably Jio Platforms' IPO execution and pricing rather than the AI build-out in isolation. Reliance's FY26 results show a business still earning the bulk of its profit from O2C, retail and core telecom — AI is a call option on future growth, not yet a reported line item investors can size precisely.
For Adani, the AI and data-centre narrative has already become more central to the group's market story. Adani Green Energy's shares rose nearly 70% over the quarter following the $100 billion announcement, with Adani Energy Solutions and Adani Power each up roughly 50%, as markets began pricing the energy units as direct AI infrastructure plays rather than conventional utilities. That re-rating cuts both ways: it raises the cost of disappointment if AdaniConneX's 5-gigawatt target slips, in a way that a similar delay at Reliance's much larger, more diversified group would likely not.
Neither company has yet disclosed a standalone AI revenue line, return-on-capital target, or payback timeline for these commitments — a reasonable omission this early in a seven-to-nine-year build, but a gap worth watching as both groups report FY27 results. Until then, the safest reading is structural rather than financial: two of India's most capital-rich private groups have concluded, independently and almost simultaneously, that owning AI infrastructure is now as strategically necessary as owning spectrum or power generation once was — and they are financing that conviction from completely different parts of their balance sheets.