Artificial intelligence may be attracting unprecedented investment, but the money flowing into the technology is still vastly greater than the revenue it generates. That imbalance could make the coming months particularly difficult for AI startups and investors, according to Nikhil Kamath, billionaire investor and co-founder of Zerodha, who has warned that the industry may be approaching a crucial financial test.
Speaking at the NDTV World Summit 2026, Kamath estimated that nearly $1 trillion is being invested in artificial intelligence annually, while enterprise AI revenue remains somewhere between $100 billion and $200 billion. By his calculations, that leaves an extraordinary funding gap of approximately $700 billion to $800 billion every year.
The figures are Kamath’s broad estimates rather than independently established industry totals, and they compare investment spending with one category of AI revenue. Nevertheless, his argument highlights a growing concern: whether the commercial returns from artificial intelligence can arrive quickly enough to justify the enormous sums being committed to its development.
Kamath questioned how long investors would remain willing to finance the difference, particularly if borrowing costs increase and capital becomes harder to obtain.
His remarks come as technology companies, venture capital firms and governments continue to make enormous bets on AI infrastructure, advanced computing, data centres and next-generation models, often on the assumption that the technology will eventually transform productivity across the global economy.
A Trillion-Dollar Investment Boom With an Uncertain Payoff
The central concern raised by Kamath is not that artificial intelligence lacks commercial potential. It is that expectations about its future economic value may have moved considerably faster than its demonstrated financial returns.
He noted that significant amounts of foreign capital, including investment from the Middle East, have already been directed toward American AI companies and infrastructure.
But maintaining that pace of investment year after year could become increasingly difficult.
The optimistic case for AI rests on its potential to reshape the services economy.
Kamath pointed to estimates placing global economic output at roughly $100 trillion to $120 trillion, with services accounting for a substantial portion. If AI can meaningfully disrupt even a fraction of that activity, the resulting productivity improvements could theoretically be worth several trillion dollars.
Such an outcome would help justify today’s extraordinary investments.
The difficulty, however, is proving that those benefits can materialise at the scale and speed investors anticipate.
Kamath acknowledged that he had not yet observed the predicted productivity transformation on the ground.
That uncertainty creates a financial dilemma. Companies must spend heavily today on infrastructure, model development, specialised chips and computing capacity, while the profits expected to justify those investments may arrive years later—or fall short of forecasts.
The result is an industry whose future depends not only on technological progress but also on investors’ willingness to continue financing that progress.
Why 2027 Could Become a Difficult Year for AI Startups
Kamath also warned that changes in global financial conditions could make the situation more challenging.
He raised the possibility of higher US Treasury yields, suggesting that borrowing costs could climb further in 2027. These were scenarios he discussed, not established forecasts.
Higher interest rates would make capital more expensive and could reduce investors’ appetite for risky, long-term technology investments.
For startups, the consequences could be particularly significant.
Unlike established technology companies with profitable businesses, many AI startups depend on repeated fundraising rounds to finance product development, computing expenses and expansion.
If investors become more cautious, companies with substantial operating costs but limited recurring revenue could face greater pressure to demonstrate commercial viability.
The concern also extends to valuations.
During periods of technological enthusiasm, investors frequently assign large valuations to businesses based on anticipated future growth. When access to capital becomes more restrictive, attention can shift toward actual revenue, sustainable margins and the ability to generate cash.
That transition can be painful for companies whose financial expectations were built around uninterrupted funding.
Kamath suggested that entrepreneurs able to raise substantial capital in the current environment and preserve it could be better positioned if fundraising conditions deteriorate.
His argument was not that AI development would necessarily collapse, but that the financial environment supporting its rapid expansion might become less forgiving.
What an AI Funding Correction Could Mean for India
For India, the potential consequences extend beyond Silicon Valley.
Indian technology startups, software companies and AI entrepreneurs are increasingly integrating artificial intelligence into products and services, while depending on global computing infrastructure, foundation models and investment flows.
A slowdown in international AI funding could affect the valuations of young companies, the availability of venture capital and the economics of building businesses around expensive AI technologies.
However, Kamath remains optimistic about India’s longer-term prospects.
During his broader summit conversation, he argued that Indian companies may find more practical opportunities in AI applications, enterprise services and cost-efficient technology solutions than in attempting to compete directly with heavily funded frontier-model developers.
He pointed to India’s engineering talent and relatively competitive labour costs as potential advantages in developing businesses that combine human expertise with AI-powered services.
That distinction may become increasingly important if investors begin demanding clearer financial returns.
Rather than competing exclusively in the expensive race to build foundational AI models, Indian businesses could concentrate on using existing models to solve specific commercial problems, improve efficiency and deliver services to international customers.
The broader question is whether AI’s economic promise can eventually catch up with its extraordinary investment requirements.
History offers numerous examples of technologies that transformed industries
while leaving many early investors with disappointing returns. The commercial importance of a technology does not necessarily guarantee that every company financing its development will succeed.
For now, Kamath’s warning underscores the uncertainty surrounding the AI investment boom.
Artificial intelligence may ultimately deliver enormous productivity improvements, transform businesses and create new industries. But the transition from technological excitement to sustainable commercial returns could be far more difficult than investors anticipate.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
