Private Equity's AI Problem Is Not Adoption. It Is Distribution.

Every fund now has an AI mandate. The portfolio companies have run their pilots, the pilots have been written up, and most of them were called a success. On paper, adoption is done.

What happens in practice is that the work carries on exactly as before.

How it happens

Not through resistance. Through handover. The mandate is issued at the fund, where the attention is. A pilot is scoped, a vendor is chosen, a sponsor is named, and for a few weeks the tool is in use. Then the sponsor goes back to the day job, the vendor moves on to the next account, and the tool stays licensed and reported upward as live, while only a handful of people ever open it.

Nobody decided to let it lapse. Every step was reasonable on the day. The result is a portfolio that has adopted AI in every board pack and distributed it almost nowhere.

Where it shows

You can report a portfolio like that for a long time. The trouble starts when somebody asks a question the pilot cannot answer.

• What did the AI decide last week, and who checked it?
• Which numbers moved because of it, and by how much?
• Who in the company owns it now the sponsor has gone?
• Would it survive a change of manager, or of vendor?

In one portco, a few of those go unanswered. Across fifteen, they become a pattern. It shows as value creation plans that list AI in every company and credit it in none, and eventually as a diligence process in which a buyer treats the whole AI story as a claim rather than as evidence.

A pilot proves the tool works. Only use proves the company changed.

Adoption was the easy part

Adoption is a decision, taken once, at the centre. Distribution is the slower work of getting the capability to the people who actually run the business: data clean enough to feed it, a process redesigned around it, and an owner on the ground whose job is to make it part of the week. A mandate cannot supply any of that. It has to be built inside each company, by people close enough to the work to know where AI belongs and where it does not.

The question for the next portfolio review

The question worth asking is not which companies have adopted AI. It is which ones can show it in use, owned and measured, without the sponsor in the room.

Everything else is secondary. A company does not need the most advanced model or the largest budget, as long as what it has is used, owned and measured the same way as everywhere else in the portfolio.

The funds that pull ahead will not be the ones that adopted first. They will be the ones that treated adoption as the start of the job, built the same three foundations into every company (clean data, a named owner and a measure agreed before launch), and made sure the capability reached the people doing the work. Value that never reaches the operators never reaches the exit, or the distributions that follow it.