 "India's AI infrastructure falls short, needs more investments: VCs")
India's funding and technological infrastructure are inadequate for making the country a global leader in the high-growth sector of artificial intelligence (AI), said venture capitalists (VCs).
The deep technology will need longer funding cycles, they said at the Capital Connect 2025 event organised by TiE Delhi-NCR, a group promoting entrepreneurship. They highlighted critical gaps in infrastructure, including limited access to graphics processing units (GPUs), insufficient data centres, and few domestic large language models (LLM).
Concerns about India’s AI readiness come amid a Chinese AI company, DeepSeek, announcing a surprising advancement that has shaken the technology industry. DeepSeek’s LLM, the foundational basis of any AI model, delivers performance comparable to OpenAI’s, Google’s or Meta’s popular offerings while requiring significantly less computing power.
Girish Shivani, executive director and fund manager of YourNest Venture Capital, said India is still taking baby steps to build AI infrastructure. “The bigger play is in the infrastructure, in managing and building AI and machine learning operations. That's where a lot of money needs to be deployed right now so that we can stay ahead of the curve over the next couple of years,” he said at the event on Friday.
The Union Cabinet in March 2024 approved the IndiaAI mission with a budget outlay of Rs 10,371.92 crore to assist AI and related technologies’ development. For 2024-25, the Electronics and IT Ministry was allocated Rs 551.75 crore for the mission.
Mohit Sadaani, managing director, DeVC & Z47, he is bullish about the technology and has invested in agentic AI, a software system or programme that does complex tasks with minimal or no human oversight. “We have about 11 investments already in the AI space,” he said.
Also Read
Sadaani said there has been sizable VC investments in AI in the United States (US) and other advanced countries, but that is not the case in India.
“I don't think we have the depth of the technology that is required today or even the infrastructure to go as deep as the US. There's a lot more money circling in that ecosystem today than here,” said Anirudh A Damani, managing director of Artha India Ventures.
India is likely to immediately focus on AI applications rather than becoming an infrastructure player. “We have seen a couple of companies do really well, and those are agent-led AI companies that are replacing work and making it faster. But I don't see India becoming an infra player in the long run,” Damani said.
Padmaja Ruparel, co-founder of IAN Group (formerly Indian Angel Network), an early-stage investing platform, said India has an opportunity in deep technology but that sector takes a long time to develop and succeed. Funds investing in this area should have more time to show results.
“We have been pushing the government to anchor thematic funding, be it deep tech or biotech. We suggest that daughter funds (that invest in MSME units) should be allowed at least 10+1 or 10+2 years because deep tech is a long gestation period. If we want to emerge as a global player in deep tech, the government can focus by giving 50 per cent of funds to daughter funds,” she said.
Shivani said the country is in the second tier with respect to GPU access. “There is clearly a mismatch between where we want to be and where we are today, and the gap will only increase,” he said.