Indian family offices and angel investors have put about $760 million into AI companies so far in 2026, nearly 50% more than a year earlier, according to Business Today’s summary of a 16 October Bloomberg report citing data from the research firm Tracxn. The investors Bloomberg names include Azim Premji’s Premji Invest and Claypond Capital, the family office of Manipal chairman Ranjan Pai.
Why it matters: India’s AI start-ups raise far less long-term money than those in the US and China, and a family office can wait longer for a result than most venture funds. That patience is a real advantage, but a handful of young companies is also a risky place for family wealth. This is a different story from Catamaran’s plan to invest more in India: that was about countries, this is about one sector.
What the $760 million figure says, and what it doesn’t
The $760 million and the growth rate appear in Business Today’s summary of the Bloomberg article; Business Standard’s republished version leaves them out. Neither gives the 2025 amount the growth is measured against, or says whether the comparison is with the same months of 2025 or the whole year.
The reports also do not explain what is counted. A database can record a family office’s own cheque, or the full size of every round it joined. The difference matters: Emergent’s $130 million round alone, in which Claypond was one of several co-leads, would be about a sixth of the total if counted in full.
The AI deals we could check
We checked the named deals against company announcements and the business press:
- Pramaana Labs. A $27 million seed round in June, led by Khosla Ventures, with Premji Invest among the other investors. Founded in 2025 by IIT Madras graduates, with offices in Bengaluru and San Francisco, it checks AI answers against written rules in fields such as tax and law.
- Emergent. A $130 million Series C in July, led by Creaegis, with Claypond Capital among the co-leads. Inc42 put the valuation at $1.5 billion. Its “vibe coding” platform builds apps from plain-language descriptions. The company says yearly revenue is running at about $120 million, a figure not publicly audited.
- Flam. A $40 million Series B in September, led by QED Investors, with Claypond Capital joining. Flam makes interactive video and 3D content with AI, and is headquartered in San Francisco with teams in India and Japan.
Bloomberg calls Premji Invest the biggest Indian family office investing in AI worldwide, without saying by what measure, and lists stakes in Databricks, Canva, Writer and Spotdraft, citing Tracxn. Premji Invest itself says it is a venture capital firm, not a family office. The reports also name QED Innovation Labs, the investment vehicle associated with CRED founder Kunal Shah, who said in June that he was joining Meta to lead WhatsApp globally. (QED Investors, which led Flam’s round, is a separate firm.)
Two patterns stand out. The family offices mostly joined rounds led by venture funds rather than leading them. And “home-grown” is a loose label: several of these companies have Indian founders and teams, but also a base in San Francisco and much of their business abroad. Inc42 reports that India provides only about 6% to 7% of Emergent’s revenue.
How big is India’s AI funding gap?
Stanford’s 2026 AI Index gives the scale. It counts venture capital and private equity rounds for AI companies that have raised more than $1.5 million, using data from Quid. From 2013 to 2025, private AI investment totalled about $757 billion in the US, about $132 billion in China and about $15.4 billion in India, eighth among the 15 countries in its chart, just behind France. In 2025 alone, the US took in about $286 billion and India about $4.1 billion. The index leaves out government guidance funds, so it likely understates China. Its figures and Tracxn’s come from different databases, so they should not be combined.
Why a family office can be patient capital
“Patient capital” is money that can wait years for a return without being forced to sell. A typical venture fund raises money from outside investors, such as pension funds, for a fixed life of about ten years, sometimes extended briefly. It must hand the cash back within that window, so it needs exits on a timetable.
A family office usually invests the family’s own money. There are no outside investors to repay and no end date, so it can hold a company longer, invest again in later rounds, or wait out a bad market. Riddhiman Jain of the wealth manager Waterfield Advisors told Bloomberg that “AI needs patient capital, which is scarce.” The same logic explains why, in BNY’s latest survey, family offices hold most of their portfolios in alternatives they cannot sell quickly (our summary).
The risks of patience
- Concentration. A family whose fortune came from technology adds to the same bet. We list concentration among five classic mistakes.
- Too few bets. Start-up returns are lopsided: most return little, a few return a great deal. A venture fund spreads money over dozens of companies; a family office with a handful of stakes can miss the winners.
- Valuation. A start-up’s value is the price of its last round, set by a small group of investors. At $1.5 billion, Emergent is valued at more than 12 times the revenue it reports. Such prices can fall sharply, and a private stake is hard to sell.
A fair caveat
The growth figure rests on one news report, seen only in summary, and one database, with no published base. The deal sizes above are for whole rounds; how much each family office put in was not disclosed. Start-up valuations are set privately and can change quickly.
Five lessons to take away
- Ask what a headline number counts. Own cheques and whole rounds give very different totals.
- Patience comes from structure. No outside investors and no fund end date are what let a family wait.
- Co-investing is not outsourcing. Joining a round led by a known fund still needs the family’s own review.
- Size each bet so you can survive losing it. In start-ups, total losses are normal.
- Look at the whole family balance sheet. A sector bet sits on top of whatever made the fortune.
Questions to bring to your advisors
How much of our wealth already depends on technology, and how much more would a start-up programme add? Can we assess a deal ourselves, or are we relying on the lead investor? What is the most we could lose without changing how the family lives? These are questions to bring to your advisors, not advice.
