Private Equity in AI

PE buys cash flow, not option value. Venture funds a lab with no revenue; PE underwrites AI-enabled software that already bills customers and can support leverage. Vista Equity Partners and Thoma Bravo built franchises on recurring-revenue software — the same asset class now being repriced as AI-enabled — while Blackstone and KKR write large structured financings.

Because AI companies now produce the cash flows PE models require. Cognition's run-rate revenue nearly doubled from US$492M to ~US$900M in the months before its US$2B Series E — recurring, forecastable revenue that makes an AI-native company underwritable through a buyout or growth-equity lens rather than as a research bet.

Control and structure. At that size PE can fund buyouts, recapitalizations and infrastructure-scale projects — KKR's Helix Digital Infrastructure targets digital infrastructure assets — where venture cheques only buy minority stakes. PE also brings operating playbooks: pricing, margins and roll-ups that convert an AI feature into durable business-model change.

Fund life and exit design. A PE vehicle expects to own for five to seven years and exits through sale or refinancing, while a crossover fund can enter a private AI round and ride it toward a listing on a much shorter clock. PE's patience is contractual — hence its preference for cash flows provable today, not promised at IPO.

Blackstone joined Coatue, MGX, T. Rowe Price and Sixth Street Growth in Databricks's US$5B strategic raise in August 2026 — infrastructure-scale capital around a revenue-generating platform. KKR's Helix Digital Infrastructure and Thoma Bravo's software take-privates mark the other entries: owning compute and cash-flowing software rather than model risk.

A buyout concerns ownership and control; growth equity may supply capital without a full acquisition. They are not interchangeable simply because the manager is a private-equity firm. Transaction type determines how a deal belongs in a financing dataset.