- Published on
Insight Partners’ Deven Parekh on why the firm is diversifying while everyone else bets the farm on OpenAI and Anthropic
- Authors

- Name
- DeepHorus
At TechCrunch's StrictlyVC event in New York on Thursday night, Insight Partners managing director Deven Parekh made a case that has become contrarian in modern venture capital: his firm is deliberately staying diversified while rivals concentrate their bets on frontier AI labs such as OpenAI and Anthropic. The comments, delivered in a candid on-stage interview, matter because they frame a widening philosophical split over where the enormous value created by artificial intelligence will ultimately settle.
Parekh has co-run Insight Partners for 26 years, and the firm now manages roughly $90 billion in assets — a scale that would normally invite loud public commentary. Yet Insight has kept a famously low profile, preferring what Parekh called letting "the portfolio do the talking." The firm has led and co-led numerous rounds in Databricks and holds stakes in both OpenAI and Anthropic, giving it exposure to the frontier while avoiding a single-minded wager.
On AI risk, Parekh aligned himself with Keith Rabois, who had dismissed researcher-driven alarmism earlier that evening. Parekh acknowledged the possibility that a non-state actor could use an open-source model to build a biological weapon, but argued the probability is higher that AI dramatically shortens drug development timelines. As an NYU Langone board member, he cited systems that scan 50 million patient records to flag a 25% heart attack risk for patients arriving for unrelated care.
Technically, Insight's approach is not a fixed allocation. Parekh described the split between early-stage, growth, buyout, and secondary investing as temporal rather than static, because the firm invests globally and adjusts by market cycle. Looking across its last seven funds, he said, an observer would find genuinely different percentages devoted to each stage — a deliberately flexible machine rather than a rigid formula.
Parekh was equally candid about losses and conflicts. He discussed deals Insight did not win, naming buzzy AI legal-tech startup Legora as one that got away, and addressed conflicts of interest inherent in venture investing. He also gently mocked the industry's new habit of instant expertise, noting that every venture capitalist now poses as an authority on epidemiology or geopolitics.
The strategic implications cut both ways. If the frontier labs deliver transformative returns, concentrated funds will capture outsized upside that diversified portfolios cannot match. But if value accrues to applications, infrastructure, and verticals like healthcare — where aging populations and scarce clinicians create durable demand — Insight's breadth may prove the safer, more compounding position over a full cycle.
Looking ahead, the debate will intensify rather than resolve. TechCrunch Disrupt 2026 will host OpenAI, Anthropic, Replit, and others across six industry stages, putting the concentration thesis on public display. Meanwhile, Insight signals it will keep deploying steadily across stages and geographies, betting that patience and portfolio breadth outlast the loudest voices in the room.