The conversation about artificial intelligence in private markets tends to happen at two unhelpful extremes. At one end, technology optimists describe a near future in which AI transforms every stage of the investment process, compresses the advantages of the largest platforms into accessible tools, and systematically improves decision quality across the industry. At the other, experienced practitioners express familiar scepticism about a technology cycle that, in their view, overpromises and underdelivers, and whose most visible effects are a proliferation of dashboards nobody reads and a generation of Associates who outsource their thinking to Claude.

Neither position is quite right, and neither is particularly useful to the investor trying to make sense of what is actually happening and what it means for how private markets will work over the next decade.

Our view is that AI is already changing specific and identifiable parts of the private equity process in ways that are real and worth understanding. It is not changing other parts, including some of the most important ones, in the ways that are sometimes claimed. And there is a third category of questions, concerning the portfolio companies that private equity manages rather than the managers themselves, where AI’s implications are more consequential, less well understood, and receiving considerably less analytical attention than they deserve.

This paper attempts to separate those three categories clearly, to be honest about the genuine uncertainty that remains, and to offer a view on what actually matters for investors and managers navigating this moment.