India plans $25 billion deep‑tech fund focused on AI and chips

A $25 billion deep‑tech pool, even if disbursed over several years, would make India one of the largest state‑backed financiers of AI and semiconductor startups outside the US and China. Strategically, it is a bid to move from being primarily a services and deployment market for imported models into owning more of the compute, model and tooling stack. If structured well, such a pool can de‑risk early‑stage bets on sovereign foundation models, domain‑specific chips and AI‑first industrial automation.

In the AGI context, this does not instantly create a new frontier lab, but it changes the medium‑term talent and capital landscape. Indian founders who might previously have relocated to Silicon Valley or London for serious frontier funding could now plausibly build at home, especially in verticals like healthcare, fintech and defence where India has large proprietary datasets. It also signals to global providers that India wants leverage in negotiations over model access, safety assurances and localisation, not just cheap API credits.

The risk is that if the pool is spread too thinly across sectors or captured by politically connected incumbents, it becomes more industrial policy than frontier push. But if even a fraction flows into genuinely novel model research, AI‑specific fabs or large‑scale evaluation infrastructure, India will be a more material player in the race to AGI than its current share of lab‑level breakthroughs suggests.