Insight Partners isn't betting the farm on OpenAI and Anthropic — and here's why

While venture funds pile into OpenAI and Anthropic, Insight Partners is deliberately staying diversified. Speaking with TechCrunch at the StrictlyVC event, Deven Parekh — who has co-led the $90 billion firm for 26 years — laid out his reasoning, which he called unusually candid for an industry where investors would rather make noise on podcasts and social media. Insight's stance is different: let the portfolio do the talking, not the managers.

Against betting on the two front-runners

In the first half of the year, OpenAI and Anthropic together took in roughly half of all venture money. Parekh admits that if a quarter of his fund were parked in Anthropic, the numbers would look better right now. But he knows funds that raise all their capital in a month on the pitch that “35–40% will go into one of these two companies,” and he considers that kind of concentration risky. Insight is now launching its thirteenth fund and thinks in terms of a ten-fund horizon, not a single lucky moment. Historically, he says, the business rewards diversification — though there are exceptions like Founders Fund and Thrive, which know how to play it concentrated.

Even so, Insight owns stakes in both OpenAI and Anthropic — something that used to be taboo. Parekh explains that at the late stage, with no board seat, it's simply buying a good stock. The internal debate was more about whether they should have gotten in earlier. The restrictions on sharing information and the refusal to invest in direct competitors apply only to the early Series A and B rounds.

The small-check strategy

Parekh is wary of the current run-up in valuations, which reminds him of 2021: rounds move so fast that there's almost no new data, and the investor pays more without reducing the risk. The logical response is to get in earlier with small checks of $20–25 million instead of $500 million, and to build up the position in the winners. That's what happened with Wiz: the Series A and follow-on investments returned far more than a one-time bet would have, while a failure would have barely dented a fund of that size. The Armis case is telling too: Sequoia grabbed the deal at first, but an Insight partner kept the relationship alive with a $5 million check, bought out the entire cap table a year and a half later, and this year sold the company to ServiceNow for $7 billion.

Not every deal gets won: the legal AI startup Legora went to General Catalyst, even though partner Jeff Horing flew to Stockholm. “The world is big; you don't need to win every deal,” Parekh says.

Liquidity beats paper multiples

The investor also stresses the value of returning money to partners. Many funds have raised large amounts of capital but returned nothing to their LPs, and for that reason won't raise a next fund. Over two years, Insight has returned more than $20 billion through sales and IPOs. His advice to founders is not to wait for endless growth: “You can't compound $40 billion at 50% every two months without becoming the world economy.” He expects that in the next 18 months SpaceX, Anthropic and OpenAI will go public — each worth more than a trillion.

Source: TechCrunch