India can catch up in AI by focusing on applications

The AI race is fast turning into a battle of deep pockets. The US is pouring billions into frontier models, advanced chips and data centres, while China is building competing capabilities. India, meanwhile, risks being left behind in the most capital-intensive phase of the tech revolution.

Does missing the first phase mean missing the entire opportunity? History suggests otherwise. India missed much of the manufacturing boom that transformed East Asia. It still built a globally competitive ITeS industry during the same period. AI could offer a similar second chance - provided India plays to its strengths.

In the 1990s-early 2000s, China and other East Asian economies strengthened their position in global manufacturing supply chains. India struggled to replicate their success. But it found an alternative: benefiting from tech advances, outsourcing and growing global demand, its services exports grew more than 80x since the 1990s, and now contribute around 55% to GVA.


Missing manufacturing was not inconsequential. India did forfeit opportunities in employment, exports and industrial capabilities. But a country can miss one structural boom and still capture value from an adjacent opportunity that aligns better with its existing strengths.

Phase 1 of the AI boom has been dominated by foundational models, high-performance computing, advanced semiconductors and data centres. These require capital, specialised research capabilities and access to cutting-edge hardware. India has made progress. But it doesn't yet command the global position enjoyed by leading US tech companies.

The centre of gravity could gradually shift - from expensive business of building ever-larger models, to the far-less-settled challenge of putting them to work across the economy, from customer service and software development to banking, healthcare, manufacturing and logistics. This is where India's existing capabilities matter. Its large technical workforce, globally integrated ITeS industry, DPI, and experience in delivering technology solutions to international clients provide a foundation for AI applications and integration.

An October 2026 World Bank report, 'India and Artificial Intelligence: Seizing the Development Opportunity', makes a similar case. It identified India's technically skilled workforce, IT ecosystem and digital infras as strengths that could give the country an edge when it comes to AI application. Private AI investment in India has risen from $1.2 bn in 2024 to $4.1 bn in 2025, while employment in GCCs has increased from 1.9 mn to 2.36 mn professionals.

But having the right ingredients may not always translate into a perfect dish. Indian firms must move beyond cost-efficient execution, and move to proprietary applications that can help integrate AI into client workflows at scale. Otherwise, AI could commoditise parts of the very ITeS industry that has the potential to anchor its next growth phase.

Alternatively, the extraordinary investment cycle could disappoint. If incremental spending on computing infra generates progressively lower returns, revenues fail to justify valuations, interest rates rise, or public resistance slows adoption, AI investment could decelerate sharply.

The recent debate around OpenAI illustrates the uncertainty. Its Sept 2026 annualised revenue was nearly $50 bn, much below a previously indicated $70 bn figure. To be sure, the difference also reflected differences in the treatment of cloud partner sales.

A correction could benefit India if it makes computing cheaper and allows domestic businesses to adopt AI without replicating the enormous infrastructure investments made elsewhere. India could avoid some direct capital losses, while capturing productivity gains from affordable technology.

But as the 2000 dotcom bust showed, investors who avoided excesses of the bubble were spared heavy losses. It, however, didn't make them immune to the economic consequences of a tech revolution reshaping entire industries. Similarly, India may escape some of the financial fallout if the AI investment boom unravels. But it can't escape the cost of falling behind in its adoption.

Ultimately, the bubble may or may not deflate. But India's fortunes from here onwards will depend on whether it can turn its existing technology ecosystem into a springboard for the next lap of the AI race.