Nikhil Kamath warns of AI bubble gap and urges Indian startups to raise capital
Zerodha co-founder flags an $800 billion gap between global AI spending and revenue while highlighting India's data centre cost advantage.

Zerodha co-founder Nikhil Kamath has advised startup founders to raise capital proactively, warning that global investment in artificial intelligence is vastly outpacing actual commercial revenues and creating market risks as borrowing costs rise.
Speaking at the NDTV World Summit and in broadcast interviews, Kamath estimated that around $1 trillion is being invested in AI globally every year, whereas enterprise adoption generates only $100 billion to $200 billion in revenue. That leaves a huge shortfall that investors must continuously bridge.
| Metric | Figure | Source / Context |
|---|---|---|
| Global annual AI investment | $1 trillion | Estimate cited by Nikhil Kamath |
| Global enterprise AI revenue | $100 billion – $200 billion | Annual adoption earnings |
| Annual AI revenue gap | $700 billion – $800 billion | Unfunded gap required from capital markets |
| India data centre setup cost | 30% – 40% cheaper | Cost per megawatt compared to the West |
| US Treasury yield cited | 5% | Benchmark for rising global borrowing costs |
Why startups should raise capital now
Kamath warned that 2027 could prove challenging for startups if investor appetite cools. With US Treasury yields around 5%, higher interest rates will make capital significantly more expensive across global markets.
Many American AI companies lack defensible competitive moats, Kamath noted, leaving their valuations vulnerable if growth and profitability projections fall short. He recommended that founders secure funds now while liquidity remains accessible, ensuring sufficient runway for coming years.
India's infrastructure advantage and local data edge
Despite valuation concerns globally, Kamath argued that India is uniquely positioned to benefit from the technology's evolution. Rather than competing strictly on raw risk capital for foundational models, India holds structural strengths in infrastructure and market context.
Setting up a megawatt of data centre capacity in India costs 30% to 40% less than in Western nations, supported by lower power costs. Kamath emphasised that value creation will increasingly move toward businesses that control local data and understand specific market contexts, rather than staying concentrated solely among US model developers.
He also supported open-source AI models, predicting that Indian enterprises will prefer keeping proprietary data inside domestic ecosystems rather than sending it to foreign platforms.
To build on these strengths, Kamath suggested India encourage greater economic competition between states, where chief ministers operate like corporate CEOs competing aggressively to attract investment, infrastructure, and technology jobs.
Key numbers
| What | Figure | Source |
|---|---|---|
| Global annual AI investment | $1 trillion | Nikhil Kamath estimate |
| Global enterprise AI annual revenue | $100 billion to $200 billion | Nikhil Kamath estimate |
| Annual AI funding gap | $700 billion to $800 billion | Nikhil Kamath estimate |
| India megawatt data centre cost advantage | 30% to 40% cheaper than Western markets | Nikhil Kamath estimate |
In this story
| Name | Role in the story |
|---|---|
| India | Nation whose AI ecosystem and data centre advantages were highlighted. |
| Nikhil Kamath | Co-founder of Zerodha and investor who commented on AI funding and startup capital. |



