Winners and Losers: How AI Is Splitting Private Equity in Two

The working assumption going into 2026 was that private equity was turning a corner. Deal activity had recovered in the second half of 2025, fundraising was stabilising, and the Iran conflict, when it came, was initially treated as one more external shock to be absorbed and priced around. Three months on, that assumption looks increasingly difficult to sustain. What is emerging instead is something more structurally interesting: a market bifurcating in real time between firms paralysed by their existing exposures and firms actively repositioning around them.

The Freeze

The retail channel, which the industry had spent years cultivating as its next major source of capital, offers the clearest window into how broad the damage has become. New commitments to US evergreen private equity and venture capital funds fell 2 per cent from the fourth quarter and were up just 2 per cent year on year, compared with a 55 per cent year on year increase at the start of last year. Private credit inflows fell even more sharply, down 30 per cent from both the fourth quarter and a year earlier. Overall alternatives fundraising is now running at a pace of $180 billion for 2026, against $211 billion in 2025, with RA Stanger chief executive Kevin Gannon warning that fundraising was on course to decline this year if first quarter trends continued¹.

The geographic picture is equally sobering. India, which had been among Asia's most attractive destinations for international buyout groups, drawing tens of billions from the likes of Blackstone and KKR, saw private equity deal value fall 33 per cent year on year to $19.6 billion in 2025². Foreign funds have pulled more than $20 billion out of Indian equities since the start of the Iran conflict, and UBS cut its GDP growth forecast for the country to 6.2 per cent, citing the energy shock and supply disruption linked to Middle East turmoil. The geographic diversification thesis, the idea that emerging markets could absorb capital cooling on developed market software assets, is itself under pressure.

But the more durable problem is not geopolitical. It is technological.

The Software Reckoning

Private equity's most productive hunting ground over the past decade has been software, specifically SaaS businesses with recurring revenues, high margins, and the kind of predictable cash flows that justified aggressive entry multiples. That investment thesis is now being fundamentally questioned. Concerns about AI's impact on software companies were a major driver of investor requests to withdraw from private credit vehicles that had lent heavily to them earlier this year. Blue Owl halted redemptions at its inaugural retail fund, with writedowns appearing at other firms.

The fear has migrated from credit into equity. EQT's Gustav Segerberg acknowledged directly that the private credit issues were bleeding into other asset classes, telling shareholders: "What's happening on the credit side of course flows into the other asset classes." EQT itself estimated it would have raised more than a billion euros more in the quarter had the private credit situation not weighed on broader sentiment¹.

The concern driving all of this is straightforward. If AI tooling can replicate the core functions of a software business (customer support automation, code generation, document processing, data analysis), then the competitive moat that justified a high purchase price may no longer hold. Private equity does not invest in narratives; it invests in cash flows and exit values. When both become difficult to model with any confidence, capital stays on the sidelines. As Prabhav Kashyap, a New Delhi based partner at Bain, put it, the overall feeling is that fundamentals remain intact but "the deployment is still being quite disciplined, people are being very selective."²

Visma: The Anatomy of a Frozen Exit

No single case illustrates the software problem more vividly than Visma, the €19 billion accounting software company owned by London based buyout group Hg. The planned IPO, which had been described as London's listing of the decade and a potential €30 billion event, is now on ice. The listing was derailed by the AI driven selloff in software stocks, a phenomenon the market has taken to calling the SaaSpocalypse, a reference to the SaaS model that investors had previously treated as the closest thing in technology to a guaranteed revenue stream³.

The Visma situation concentrates several of the industry's most acute problems into one asset. Hg's portfolio, as represented by its London listed investment trust, carries average net debt of more than seven times earnings and valuations at 25 times earnings as of December, with the trust's shares falling 26 per cent since the start of the year. A $5 billion fund holding part of the Visma stake carries a loan equal to about a fifth of its value. Jefferies analyst Matthew Hose warned that in a polarised environment of AI winners and losers, leverage at the fund level could cause difficulties in specific portfolio companies to cascade.

The valuation question is the most exposed. A 2023 deal valued Visma at €19 billion, with advisers last year suggesting it could reach €30 billion at IPO. It is now described as worth "more than €20 billion", equivalent to 20 times adjusted ebitda, at a time when comparable public market software groups trade at around 10 times forward earnings. The gap between private and public market valuations for software assets is not a new observation, but Visma makes it concrete in a way that is hard to dismiss.

Defenders of Visma point to its 20 per cent year on year revenue growth, relatively low leverage at the company level, and the proprietary data advantages that give it an edge in deploying AI internally. The argument is that complexity and regulatory specificity (Visma handles payroll, accounting, and tax compliance for small businesses across 28 countries) provide a degree of protection that simpler software models lack. Some Hg backers told the FT they were not worried and trusted Hg's approach to AI deployment ahead of peers. Others conceded they could not predict how any business would fare given the pace of AI development³.

That honest uncertainty is itself the problem. In an asset class that prices future cash flows, "we cannot predict the outcome" is not a foundation for an IPO.

The Other Side of the Bifurcation

What makes the current moment analytically interesting is that not all firms are frozen. The same period that saw Hg's Visma IPO remain on ice also saw EQT make a £10.6 billion takeover approach for Intertek, the FTSE 100 testing and quality assurance business, at a premium of up to 62 per cent to its share price before EQT's approach⁴. Intertek's board signalled it was minded to recommend the offer.

The contrast is instructive. Intertek has nothing to do with software. It operates in physical supply chain verification, product testing, and risk management, areas where AI disruption is far less acute and where depressed UK public market valuations have created a genuine pricing opportunity. The deal is the latest in a series of private equity buyouts of London listed businesses, as depressed UK equity valuations attract buyout capital at the same moment public market investors have pulled back⁴.

EQT's move is not isolated. Infrastructure and real estate vehicles have weathered the current slowdown considerably better than software exposed funds, with quarterly inflows to evergreen property funds up more than 25 per cent year on year. KKR, Ares, and Brookfield all reported stronger fundraising and inflows in both asset classes, with Ares telling shareholders it had experienced accelerating demand outside of US private credit. The firms gaining ground are those that either pivoted away from software before the selloff, or are large and diversified enough to redeploy capital into sectors where the AI overhang does not apply.

My view is that this bifurcation will deepen before it resolves. The firms that went deepest into SaaS between 2019 and 2022, at peak multiples and with leverage, are now holding assets that are simultaneously harder to value and harder to exit. The retail capital they were counting on is retreating. The geographic alternatives are under their own pressures. And the AI disruption repricing their existing portfolios is, paradoxically, creating new investment opportunities in infrastructure, compute, and enterprise integration that they are not yet positioned to capture.

The firms that navigate this successfully will not be the ones that waited for conditions to normalise. They will be the ones that made a clear eyed decision about which side of the AI divide their portfolio sat on, and acted on it. EQT buying Intertek while simultaneously warning about zombie funds and the SaaSpocalypse is, whatever else one thinks of it, a firm that has made that call.

Footnotes

1. Alexandra Heal and Eric Platt, Retail investors cool on private equity as jitters spread↗ (opens in a new tab), Financial Times, 10 May 2026. ↩ ↩²

2. Chris Kay and Krishn Kaushik, Private equity cools on India amid lofty prices and economic jitters↗ (opens in a new tab), Financial Times, 14 May 2026. ↩ ↩²

3. Alexandra Heal and Antoine Gara, Can Britain's star tech investor dodge the SaaSpocalypse?↗ (opens in a new tab), Financial Times, 6 May 2026. ↩ ↩²

4. Aaron Kirchfeld, Intertek set to agree £10.6bn takeover by private equity group EQT↗ (opens in a new tab), Financial Times, 13 May 2026. ↩ ↩²