India's AI Spend Is Accelerating. Its Data Foundations Are Not.
Indian enterprises are deploying AI faster than most other countries. They are also struggling to show for it. The problem is not ambition. It is the foundations.
Image Credit- Chetan Jha/ MIT Sloan Management Review India
Key Takeaways
01
Indian enterprises are spending more on technology than global peers and deploying AI at above-average rates, yet only 15% of business leaders view IT as truly strategic. The gap is not ambition. It is foundations: 90% of enterprises lack sufficient data infrastructure for enterprise-wide AI scale.
02
India’s venture market is following the global shift toward concentration, but more selectively than the US. Seed-stage funding dropped 30% in 2025. Late-stage fell 26%. Capital is consolidating around firms that demonstrate product-market fit and sustainable unit economics.
03
The priority for Indian technology leaders in 2026 is not launching more pilots. It is closing the pilot-to-production gap. Globally, 88% of AI proofs of concept never reach production. India’s high deployment rate does not exempt it from this structural failure mode.
The first phase of the global AI boom rewarded speed. Companies launched products before anyone had worked out who would pay for them or why. Being present in the market mattered more than building something sustainable.
India entered that phase later and at lower capital intensity than Silicon Valley. Now the same reckoning is arriving, in its own form. Indian enterprises are spending more on technology than their global peers, deploying AI at higher rates than in most comparable economies, yet still failing to demonstrate strategic returns at scale.
Bain and Company’s India Enterprise Technology Report 2026, published in May after surveying more than 250 chief information officers (CIOs), chief digital officers, and senior technology executives, puts the gap in precise terms.
Indian enterprises are spending 150 to 200 basis points more of revenue on information technology than their global counterparts. IT spending is projected to grow 6 to 8% in 2026, against a global rate of 4 to 6%. AI and data transformation alone are expected to account for 40-45% of change-related technology budgets this year.
Yet only 15% of the business leaders surveyed view IT as truly strategic. About 90% say their current data foundations are not sufficient to support enterprise-wide AI at scale.
More spending, deployed on inadequate foundations, produces expensive noise rather than a competitive advantage. India’s AI discipline problem is not that firms are experimenting too widely. It is that they are scaling before the ground beneath their systems is ready.
“The era of broad experimentation is over,” says Khadim Batti, Chief Executive and Co-Founder of Whatfix, an enterprise digital adoption platform. “Companies are now focused on using AI to reshape their core business.”
That focus is visible in how capital is moving. But in India, the more instructive signal is not where the money is going. It is what the money is hitting when it arrives.
India Is Deploying AI Faster Than It Is Preparing for AI
Deloitte’s State of AI in the Enterprise report for 2026, based on a survey of 3,235 senior leaders across 24 countries conducted in late 2025, found that 40% of Indian respondents reported significant or full use of AI across their organizations, compared with a global average of 28%. In product development, strategy and operations, and marketing and sales, India leads global peers in at-scale deployment.
This is not a country lagging. It is a country outpacing its own infrastructure.
The EY AIdea of India report for 2026 found that 47% of Indian enterprises now operate multiple generative AI use cases, and nearly half report that more than 21% of their proofs of concept have progressed to production. That conversion rate is higher than it was two years ago. It is still far below what the volume of experimentation would justify. IDC research found that globally, 88% of AI proofs of concept never reach production deployment. For every 33 pilots a company launches, only four reach operational use.
The bottleneck is not model quality. The models work. The bottleneck is the infrastructure beneath them.
“AI demos are cheap to build,” says Mohith Mohan, Chief Executive of MOAR Advisory. “But AI businesses are expensive to run.” Turning a prototype into a production system requires sustained investment in data pipelines, compute infrastructure, compliance architecture, and integration with the operational systems it is meant to improve.
The Bain report found that 72% of Indian CIOs cite legacy technology debt as the top barrier to transformation. Forty-nine percent point to unproven return on investment from new technology initiatives. These are not emerging concerns but structural conditions that have persisted through 18 months of accelerating AI spend, and the spending has not resolved them.
“Vision alone is no longer sufficient. AI needs to show clear business value to survive.” — Mohith Mohan, Chief Executive, MOAR Advisory |
The India-specific dimension of this is the pace mismatch. Indian enterprises adopted AI workflows faster than their global peers and are now discovering the gap between workflow adoption and enterprise transformation earlier than most.
Rajesh Nambiar, President of Nasscom, described the current moment as the ecosystem entering “a more disciplined phase” in which AI is emerging as “core infrastructure for India’s next innovation cycle.” Infrastructure, unlike a feature rollout, requires foundations. And foundations take longer to build than deployments take to launch.
Capital Is Concentrating. The Middle Is Being Squeezed Out.
The global pattern of AI capital concentration is well documented. In the first quarter of 2026, about 80% of global venture funding flowed into AI, with a handful of firms capturing most of it. According to Crunchbase, OpenAI, Anthropic, xAI, and Waymo raised $188 billion between them in a single quarter. Stanford University’s AI Index put enterprise AI infrastructure spending at $37 billion in 2024. Gartner projects global AI spending will reach $2.52 trillion in 2026.
India is not exempt from this concentration, but its version looks different. In 2025, India’s startup ecosystem raised $10.5 billion, a 17% decline from 2024, according to Tracxn data cited by TechCrunch. The number of funding rounds fell 39%, to 1,518 deals. Seed-stage funding dropped 30% as investors pulled back from experimental bets. Late-stage funding fell 26% as scrutiny of scale and profitability intensified. AI startups in India raised just over $643 million across 100 deals in 2025, a modest 4% increase from the year before.
The pattern in India is selective concentration, not the mass capital injection visible in the US. Investors are writing fewer, larger checks into companies that can demonstrate product-market fit and a path to sustainable unit economics. The deep-tech funding surge of 37% in 2025, driven heavily by AI, reflects this: money is moving toward commercially viable AI, not toward experimentation.
“Companies are becoming more selective about where they put their money and talent,” says Mohan. “Many of these projects were not failures. They just did not show enough demand or long-term value to justify continued investment.”
For India’s mid-market software firms, this creates a specific pressure. AlixPartners projected a 30 to 40% surge in enterprise software mergers and acquisitions globally in 2026 as AI disruption forces mid-market firms to consolidate or exit. In India, that pressure is compounding the existing challenge of proving AI returns on constrained data foundations.
“This is not the end of experimentation,” says Raghu Pareddy, Chief Executive of Wissen Technology. “It is becoming more focused. Companies are putting their energy into ideas that can scale and deliver real business value.” That is the right instinct. The question is whether the foundations required for that scale are being built at the same pace as the ambition.
Research Context
Primary interviews with Mohith Mohan (MOAR Advisory), Raghu Pareddy (Wissen Technology), and Khadim Batti (Whatfix). Quantitative sources: Bain India Enterprise Technology Report 2026 (May 2026); Deloitte State of AI in the Enterprise India 2026; EY AIdea of India 2026; Tracxn/TechCrunch India startup funding data 2025; IDC AI POC-to-production research; Stanford AI Index 2025; Gartner AI spending forecast 2026; Crunchbase Q1 2026 venture data; AlixPartners Enterprise Software Technology Predictions 2026. |
What Leaders in Each Role Must Do Differently
C-Suite: Sequence the Investment, Not Just the Ambition
The Bain data is a diagnostic. Ninety percent of Indian enterprises lack sufficient data foundations for enterprise-wide AI scale. Before the next AI initiative is approved, leadership teams need to answer a prior question: is the data infrastructure beneath it capable of supporting production deployment, or is the organization about to spend capital building on foundations that cannot hold it? AI investment is now a capital allocation decision. The projects most likely to justify their cost are those that include data foundation work as a first-phase requirement rather than an afterthought.
Technology Leaders: Close the Pilot-to-Production Gap Deliberately
Globally, 88% of AI proofs of concept never reach production. The IDC finding points to a specific cause: low organizational readiness in data, processes, and infrastructure, not model failure. For CIOs and technology leaders in India, the priority in the next twelve months is not launching more pilots. It is developing the conversion discipline to take a smaller number of pilots through to operational deployment. Sixty percent of Indian CIOs plan to prioritize high-impact AI roadmaps alongside application rationalization and data modernization in the next year, according to Bain. Application rationalization is the right instinct: fewer systems, cleaner data, higher conversion rates.
Boards and Investors: Separate Deployment from Value
India leads global peers in at-scale AI deployment. It does not demonstrate strategic value from that deployment. Boards governing technology-intensive enterprises need to insist on outcome-based reporting, not implementation reporting. The question is not how many AI use cases are live. It is how many are producing measurable impact on growth, efficiency, or profitability. Bain found that enterprises with a deliberate approach to AI foundations and operating model redesign have the potential to achieve an absolute earnings improvement of 15 to 20%. That range is not available to organizations that deploy AI on top of legacy data infrastructure without addressing the foundations first.
The global AI discipline phase is about consolidation: fewer standalone products, more integration into platforms, sharper focus on returns. In India, that discipline is arriving in a different form. The challenge is not too many standalone products. It is too much deployment on too little foundation.
Indian enterprises are not behind. In several measures of AI adoption they are ahead. The risk is that scale without foundation produces a new generation of technical debt rather than the competitive advantage the investment was meant to create.
“There is a risk that companies become too cautious,” Pareddy says. “The challenge is to stay bold while still being disciplined.”
In India’s case, discipline does not mean doing less. It means building the ground that the ambition requires before the ambition outpaces the ground.
Editor’s Note: MIT Sloan Management Review’s AI Research Forum will make its India debut in Bengaluru on 23 July, bringing together enterprise leaders, researchers, and practitioners to examine how autonomous AI is moving from experimentation to governed deployment at scale. To speak, partner, or attend, register here.
