The missing half of India's tech policy

As the government deploys public capital into frontier technologies, its expert investment committees need two improvements to deliver the intended outcomes

India’s policy debate is often dominated by a single question: What should the government do? We argue about the right industrial policy, the right approach to artificial intelligence (AI), the right incentives for deep tech, and the right financing architecture for the energy transition. These are all valid questions, but they address only half of the outcome. The other half is whether we have institutions capable of implementing those policies with competence and continuity.

 

This matters most when the government is allocating risk capital. Public systems are rightly built around transparency, accountability, and process. Investment decisions require something further: Specialised domain knowledge, judgement about people and technologies, and a tolerance for uncertain outcomes. A file can establish compliance. It cannot tell you whether a first-time fund manager has the judgement to back the next generation of Indian technology companies. That capability sits in the market, with people who have backed such teams and lost money on some of them.

 

We confronted exactly this problem when I was minister of state for finance. In 2015, the government launched the India Aspiration Fund through the Small Industries Development Bank of India (Sidbi) with a corpus of ₹2,000 crore. Rather than ask officials alone to select venture funds, Sidbi constituted a Venture Capital Investment Committee drawn from India’s investment and entrepreneurial ecosystem. That committee made the investment calls, and its recommendations then went through Sidbi’s formal governance process for approval. The institution supplied the safeguards and the audit trail. The committee supplied the judgement.

 

Expert investment committees offer three advantages. First, detailed domain expertise. Evaluating a venture fund, a semiconductor company, a biotechnology platform, or an AI model requires technical and commercial understanding that administrative experience alone does not produce. Second, networks. Investment decisions are ultimately decisions about teams. Experienced investors know the founders, the technologists, and the previous limited partners, and can triangulate a reputation in a few phone calls. Third, continuity. Officers in ex-officio positions move on after two or three years. An expert can serve five years or more, which matters when funds run for 10 to 12 years.

 

The architecture was then scaled through the ₹10,000 crore fund of funds for startups approved in 2016. By December 2025, the alternative investment funds it backed had invested about ₹25,548 crore in 1,371 startups. Anchor commitments of this kind helped catalyse domestic venture capital by giving first-time Indian managers the confidence and capital for a first close, and by pulling far larger pools of private money alongside.

 

This institutional innovation has since spread. Startup India Fund of Funds 2.0, notified in April 2026 with another ₹10,000 crore, allocates across deep tech, micro venture capital, technology-led manufacturing, and agnostic funds, with a Venture Capital Investment Committee. The ₹1 trillion Research, Development and Innovation Scheme uses a two-tier architecture in which second-level fund managers evaluate projects through investment committees of independent external experts. Together, these mechanisms are creating a new institutional architecture for financing deep tech.

 

The same approach is now being considered for a proposed National Frontier AI and Compute Fund, with a reported government anchor of ₹15,000 crore to ₹20,000 crore, roughly equal private capital, and investment decisions taken by a committee of technology, investment, infrastructure, and finance experts. A graphics processing unit (GPU) cluster is capital-intensive and depreciates on a three- to four-year cycle, so capacity built against demand that does not arrive gets stranded quickly. Whether a specific cluster, with specific offtake commitments at specific prices, should be funded is an expert call.

 

Two improvements are now required. The first is to globalise these committees. India competes for capital, technology, and talent globally, and a committee evaluating frontier AI, quantum computing, or advanced materials should include people who have built companies and investment platforms in Silicon Valley, Boston, London, and Singapore. The diaspora gives us an extraordinary pool of talent that can contribute to national institution building.

 

The second is that active VC investors should generally not sit on committees allocating public capital in sectors where they are currently deploying. The recent discussion around the RDI Fund showed why perceptions matter even where disclosure and recusal rules exist. The government defended its safeguards and signalled openness to more, which is the right instinct. For public capital, avoiding actual conflicts is not enough; the institutional design must also minimise the appearance of conflict. India now has a deep bench of retired investors, former technology executives, academics, and global experts who can supply sophisticated judgement without creating avoidable conflicts.

 

Countries develop because they build institutions that decide well repeatedly, transparently, and over decades. Expert investment committees are one such institutional innovation, and they have earned their place. Improving them further is how we convert public capital into private capability, and policy ambition into Viksit Bharat.

The writer is president of the Everstone Group and a visiting professor in practice at the London School of Economics. He is a former minister of state for finance and civil aviation, and a former chairman of the Parliamentary Standing Committee on Finance