Private Equity in the Age of AI: From Cost Optimisation to Business Model Reinvention

Two years ago, the question in private equity was how AI could improve EBITDA. Today the more important question is which portfolio companies AI will make more valuable, and which it will quietly make obsolete.
In 2024 I wrote that AI could become one of the most powerful levers for EBITDA improvement in private equity portfolios. I still believe that. But the conversation in 2026 needs to move beyond isolated productivity projects.
What has changed is the depth of the impact. AI is no longer just automating tasks inside an existing business model. It is starting to change the business models themselves: how companies acquire customers, how they deliver services, how they price, and how much labour and capital they need to grow.
The two sided nature of AI in a portfolio
For a PE fund, AI now has two faces.
On one side it is a value creation lever. Customer service, finance, procurement, software development, sales operations and document heavy back office processes can be redesigned with agents and generative AI. In the best cases I have seen, this means not a few percent of savings but a structural change in the cost to serve.
On the other side it is a source of disruption risk. Business services companies that sell hours, software companies that sell seats, outsourcing providers, content and information businesses, and many professional services models face the possibility that AI will compress their pricing or allow customers to do the work themselves. Listed markets have already started to reprice parts of the software and services universe on exactly this fear.
A portfolio that looks diversified across sectors may in fact be concentrated in a single risk: exposure to labour intensive revenue models that AI can replicate.
Due diligence must change
This has direct implications for how funds evaluate new acquisitions. A traditional commercial and operational due diligence asks about market growth, competitive position, customer concentration and cost structure. In 2026 that is no longer enough.
When I conduct technology due diligence today, I add a set of questions that would have seemed exotic three years ago. How much of the company’s revenue is linked to activities that AI can perform at a fraction of today’s cost? Does the company own proprietary data, workflows or customer relationships that become more valuable with AI, or does it mainly sell access and labour? Is the technology architecture ready for AI, or will years of remediation be needed before anything meaningful can happen? Does the management team have the capability and the will to lead a transformation, or are they defending the existing model?
And perhaps the most important question of all: in five years, will this company be a beneficiary of AI adoption in its industry, or a casualty?
From projects to thesis
The funds that are getting this right are no longer treating AI as a project inside the value creation plan. They are making it part of the investment thesis itself. The entry price reflects an explicit view on AI risk and opportunity. The 100 day plan includes AI diagnostics. The operating team includes people who have actually led AI transformations, not just advised on them. And the exit narrative is built around demonstrated AI capability, because buyers will increasingly pay for it and discount its absence.
The organisational truth
In my experience, the hardest part is not technology. It is the organisation. A portfolio company that has spent decades optimising a labour intensive model will not reinvent itself because the board approved a strategy. It needs leadership, new skills, new incentives, and often a willingness to cannibalise its own revenue before a competitor does.
That is exactly where private equity, with its concentrated ownership, active governance and finite holding periods, can have an advantage over public companies. Owners who can make decisions quickly may be better placed than anyone to force the change.
Investment question: Should AI transformation become a standard component of the investment thesis for every Private Equity acquisition?


