India aims at becoming the world’s next AI superpower, but it is skipping the first step: basic research. The gap is not about ambition, but about how the nation funds its innovation, according to Amit Sheth, a professor at the University of South Carolina and one of the world’s most cited computer scientists.

Technology is measured by what are called Technology Readiness Levels (TRLs). TRL 1–3 is basic research, TRL 4–7 is applied research, and TRL 7–9 is when products are ready for deployment. 

“For every 10 or 15 papers, India has one,” he said. Sheth himself has one of the highest H-indexes in the field but admits he achieved that in the US ecosystem, not in India. Early research can take five or more years, but that is exactly where India invests the least.

“Indian venture capital has only chosen to fund [TRL] 7 to 9. Indian industry has only chosen to do product tinkering,” Sheth said, adding that at top AI conferences like AAAI, India barely shows up.

India’s Blind Spot in Funding

The government’s biggest research initiative in years, the ₹1 lakh crore Research Development and Innovation (RDI) scheme, cleared by the Union Cabinet this year in July, will only support projects that have already crossed TRL 4. 

That means fundamental research, concept validation and early lab experiments — TRL 1 to 3 — are excluded. The policy states it outright: funding is “intended only for projects that have achieved TRL-4 and above.”

This is a structural flaw. Breakthrough technologies almost always begin at the lowest levels of readiness. By cutting them out, India is shutting the door on the very innovations it claims to want.

The Technology Development Board (TDB) takes a similar approach, backing projects only in TRL 7–9. This has created systemic neglect of foundational research. The Civil Services Daily noted that the TRL-4 rule “narrows the innovation pipeline” and blocks “diverse and disruptive innovations from entering the ecosystem.”

India’s R&D spending— at 0.64% of GDP—is not only weak, but far behind advanced economies. Whatever money does exist tends to flow towards late-stage development.

The Talent Problem

Sheth argued that the funding gap leads to India’s shortage of deep research talent. Compare India with the US and China. A startup like Perplexity can raise billions in one round, while in India the typical ticket size is just a couple of million dollars. 

The scale of funding and the pool of PhDs are both missing.

He pointed to Google’s early years: its first 10,000 employees were mostly PhDs. “I want to do Google here. Where are we going to get that talent?” he asked.

Sheth trains PhDs in the US who graduate with dozens of publications, citations in the thousands, and even funding in their own names before finishing. “Compare that profile with a typical PhD we produce,” he said. The difference, he argued, is not just about curriculum but the entire ecosystem of research, mentorship, and industry connection.

India has expanded institutions like IITs and IISc, but most still prioritise teaching over research. “A great professor should not only be successful themselves but also train students so well that they can be equally successful. That has not happened enough in India,” Sheth said.

A Missed Opportunity

The neglect of TRL 1–3 hurts deep tech most. Venture data shows 80% of Indian deep tech deals are seed-stage, with very few making it to Series B and beyond. For early-stage technologies, the only funding usually comes from government grants. 

With the RDI scheme now locked at higher levels, that pipeline is drying up.

The consequences are predictable. The internet and GPS began as low-TRL, risky military projects. While failing to fund that stage, India is ensuring it misses similar breakthroughs. Universities have little incentive to pursue basic science, widening the gap between academia and industry. Talented researchers leave the country because their work won’t get supported here.

Some programs do exist. The Science and Engineering Research Board (SERB), BIRAC and NIDHI offer small grants for early research. BIRAC’s Biotechnology Ignition Grant has had a real impact. But funding levels haven’t grown in a decade, making them irrelevant compared to the scale of the RDI scheme.

Where India Should Focus

Sheth doesn’t think India should chase large language models. “Don’t go there. We can’t compete with Meta putting $100 million into a startup,” he said. Instead, India should focus on enterprise AI—using data, knowledge, and human expertise to solve specific problems in manufacturing, health, semiconductors, and other sectors. 

Indic language AI is important, but more as a public good than a global business.

For building something global, Sheth argues that the timing may now work in India’s favour. US research funding is shrinking. The National Science Foundation and National Institutes of Health have faced cuts of nearly 50%. 

“If you are not getting funding, you cannot attract top students. In the US, I spend $300,000 training each PhD, all from my own grants. If I don’t get that funding, I cannot hire PhDs, then I cannot do great research,” Sheth said.

Fixing the System

Sheth said India needs a program on the scale of China’s “Thousand Talents,” which gives returning researchers million-dollar grants to build labs. He pitched a similar idea to Prime Minister Narendra Modi years ago when he was Gujarat’s chief minister: a world-class research university focused entirely on research, not just teaching.

Building a Stanford, MIT, or Harvard in India is not a task that can be easily achieved. Nalanda 2.0 was an initiative started by Shailendra “Shail” Kumar, former president of the IIT Foundation and the author of Building Golden India. The aim was to build a university of the future, called Ekagrid, in Bangalore. 

Unfortunately, Kumar announced that the board of directors have decided to close the university initiative. 

Read: The Nalanda 2.0 Dream Crumbles

Without real investment in TRL 1–3, without building a research culture that produces world-class PhDs, the country will remain stuck at the level of “product tinkering” while others build the Googles and OpenAIs of the future. 

Or as Sheth put it: “Google could do it because it had Stanford and Berkeley. If we want to do it here, we need institutions that can produce that kind of talent.”

The post The Blind Spot of Indian Investment is TRL 1-3 appeared first on Analytics India Magazine.