India’s NSE CEO: Winning in AI doesn’t require the biggest models

By Adam Mills

AI Editor

See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.

The global AI race is often measured by who owns the largest models, the most advanced chips and the biggest data centers. But Shri Ashishkumar Chauhan, CEO of the National Stock Exchange of India, believes that scoreboard may not decide who ultimately captures the economic gains.

“You don’t have to be inventing something to be the best users,” Chauhan said during Newsweek’s July 27 “AI Impact Forum” webinar, “Is India on the Right Side of the AI Trade?” hosted by Dr. Ranjit Tinaikar.

Read More on AI

Chauhan argued that India could build an advantage by applying AI across businesses, services and everyday devices, even when much of the underlying technology was developed elsewhere.

Dr. Ranjit Tinaikar, host of Newsweek’s “AI Impact Forum,” left, spoke with Shri Ashishkumar Chauhan, CEO of the National Stock Exchange of India, right, about the country’s workforce, capital markets and position in the global AI economy.

...

Tinaikar raised a warning from Bernstein, a global equity research and brokerage firm, that India could become a “permanent customer” of AI models developed in the United States or China. Chauhan answered by looking back at India’s earlier technology history.

India did not create computer chips, operating systems, relational databases or programming languages, Chauhan noted. Its workers still learned to use those technologies and built a global industry around solving problems for companies elsewhere.

He sees another opening in AI. Open models can lower the cost of experimentation, while smaller systems can be adapted for phones, vehicles, appliances, industrial equipment and other uses that do not require the largest frontier models.

“The battle of LLMs is over,” Chauhan said, recalling a view he shared publicly in 2025. “Mostly open source will win.”

Model developers and infrastructure providers currently command much of the investment and attention. Chauhan expects more economic value to move toward the companies and countries that can apply AI to real work at an affordable cost.

India enters that competition with millions of employees across technology services firms and global capability centers. Many already build and support systems for companies around the world, giving them experience with the business processes that AI is beginning to change.

“AI will allow you to solve even larger problems at a lower cost,” Chauhan said.

From smaller models to semiconductor capacity and energy, India’s AI prospects extend across the technology stack and into everyday applications.

...

Smaller models could put more of those uses within reach. Chauhan expects more AI processing to move onto phones and other devices as computing becomes cheaper and systems become easier to customize. Open technology would allow developers to adapt models to local needs without routing every task through a centralized platform.

Large models would still serve demanding use cases, while smaller and specialized systems handle more routine work. Countries that use both effectively may capture more value than those focused mainly on building the largest systems.

Chauhan also challenged the way innovation is measured. When Tinaikar pointed to India’s comparatively low research and development spending, Chauhan argued that accounting figures do not always reveal where the technical work takes place.

Engineers in India often conduct research for American and European companies even when the spending and intellectual property appear on corporate books elsewhere, he said. Some later leave those companies and build businesses of their own.

“India is what I call a bottom-up framework,” Chauhan said.

Entrepreneurs in that system respond to commercial opportunities as they emerge instead of waiting for a national plan to direct capital into a selected industry. Chauhan said their innovations need to remain “frugal” while still being “world-class.”

Physical infrastructure will also shape how widely AI can be deployed. Chauhan said India is adding semiconductor manufacturing capacity, expanding renewable energy and exploring nuclear technologies as data centers and other computing facilities increase demand for electricity.

He also warned against assuming today’s shortages and valuations will last. Heavy spending on data centers and semiconductor production could eventually create excess capacity if AI revenues fall short of expectations.

Shri Ashishkumar Chauhan, managing director and CEO of the National Stock Exchange of India, said countries can prosper from AI without creating every underlying technology: “You don’t have to be inventing something to be the best users.”

...

The same skepticism shapes his view of closed-model companies. Chauhan doubts that a small group of providers will retain nearly all of the value created by AI, particularly as open-source and open-weight systems compete on cost and give developers more freedom to modify the technology.

“It’s going to be a garden with many flowers of different types and not one single type of flowers,” he said.

A wider mix of models could play to India’s strengths. The country does not need to lead in models, chips and applications alike to create successful companies, expand its technology workforce or develop products for global markets. Its strongest route may run through the areas where its technical skills, costs and experience already offer an advantage.

Chauhan’s emphasis on frugality also shaped his advice to AI entrepreneurs. He urged founders to keep expenses below income so they have more time to build, experiment and eventually expand.

“Who is the richest person on the earth? The person who spends less than his income,” he said.

Newsweek’s next AI webinar turns to another part of the adoption story: the customer experience. Registration is still open for “Humanizing AI: Personalization, Engagement and the Customer Experience,” taking place on Thursday, July 30, at 2 p.m. Eastern. Presented by Cognizant, the live discussion will examine where AI can improve customer service, engagement and personalization, how human handoffs should work and what companies need to address in their data, workflows and teams.

Gabriel Snyder, Newsweek’s executive editor, enterprise, will moderate a panel featuring Nick Mehta, entrepreneur in residence at Bessemer Venture Partners and former CEO of Gainsight; Arun Chandra, chief operating officer at NiCE; Jon Hyman, co-founder and chief technology officer at Braze; and Adrian McDermott, chief technology officer at Zendesk.