Private equity stumbles into AI's blind spot

Private equity stumbles into AI's blind spot

Private equity firms are confronting a fundamental crisis: they cannot reliably predict what their portfolio companies will be worth three or four years from now, when artificial intelligence keeps rewriting the rulebook every few months.

The problem runs deeper than the software holdings that took a valuation beating in recent quarters. Executives across industries told Axios this week at the Milken Global Conference that AI has shattered the modeling assumptions underlying nearly every deal structure, whether the target operates in tech, healthcare, manufacturing, or services.

The timing creates a peculiar trap. ChatGPT launched just three and a half years ago. For private equity, which typically locks in capital for three to four years minimum before an exit, that horizon now encompasses an era of radical technological flux. Claude, Gemini, and successive breakthroughs have already disrupted sectors no one predicted would be touched.

Exit multiples, the financial linchpin of any deal, once relied on educated guesses based on historical patterns and sector trends. Today, according to one veteran dealmaker, the exercise feels like throwing darts blindfolded.

Even industries that seem positioned to benefit from AI rather than be displaced by it face murky forecasting problems. The known unknowns have multiplied. Anyone claiming high confidence in the market landscape years out is either fibbing or delusional, even with backing from leading AI labs.

The paradox: private equity has more dry powder than ever. Limited partners continue writing checks despite persistent DPI headwinds. Capital will deploy, deals will close, and new portfolio companies will join the fold.

But the industry's greatest historical advantage, its long-term hold structure and patient capital, has become a liability. Three or four years is no longer a manageable planning window. It is a bet on an unknowable future.

Author James Rodriguez: "PE's entire business model is built on seeing around corners, but nobody can see anything past the AI fog right now."

Comments