Prominent private credit investors are warning that artificial intelligence is forcing a significant reckoning across the software sector. 

But they are pushing back against the most catastrophic interpretations, arguing the disruption will produce a K-shaped outcome of winners and losers rather than a broad collapse of software-as-a-service businesses.

Credit specialists at major firms including Ares, Man Group, and HarbourVest outlined their thinking at the SuperReturn International private equity conference in Berlin this week, offering a nuanced view of how AI is reshaping one of private credit’s most significant long-term investment bets.

Private Credit Has Enormous Exposure to Software

The stakes of getting this call right are significant.

Alternative lenders’ exposure to the software sector has grown to between 20% and 30% of private credit portfolios on average over the past five years, according to Kevin Marchetti, chief investment officer and head of U.S. direct lending at Man Group.

That concentration made the “SaaSpocalypse” fears that emerged earlier this year, when concerns spread that AI could make much of the existing software landscape obsolete, a particularly alarming scenario for private credit managers.

Ares co-president Blair Jacobson said his firm took notice of the debate but was not caught off guard.

“It wasn’t a big surprise,” Jacobson said. “Some companies will be able to adapt to AI, and some will face more challenges.”

Software stocks have partially recovered from their February sell-off. 

The iShares Expanded Tech-Software Sector ETF surged 21% in May and has gained 9% on a three-month basis, reflecting a market that has moved from broad sector panic toward a more selective assessment of which companies can adapt.

Mission-Critical Software Has a Layer of Protection

Ares is focusing its software lending on companies that provide mission-critical services, particularly in regulated industries where the cost of switching providers, or of failure, is prohibitively high.

Jacobson said enterprise resource planning systems embedded in regulated business processes represent a category where strong software companies retain meaningful pricing power and customer stickiness even as AI disrupts the broader market.

“The cost of failure is very high, so we think we’re going to see more bifurcated outcomes,” Jacobson said. 

He added that from a credit investor’s perspective, the current environment is actually creating attractive opportunities: spreads are widening, documentation standards are tightening, and loan-to-value ratios are declining, all of which improve the risk-reward profile for disciplined lenders willing to continue extending credit selectively.

Man Group Shifts Toward Old Economy Businesses

Man Group’s Marchetti said his firm is increasingly pivoting its direct lending focus toward what he described as old economy businesses, where AI is more likely to serve as an operational enhancer than a fundamental disruptor of the business model.

That category includes critical business-to-business healthcare services and distribution companies serving Main Street businesses, sectors where AI can drive efficiency and productivity gains without threatening the core value proposition of the business.

“The software credit thesis is being tested, but I think there are still very good companies in that space, and it’s just understanding how that will play out over the next three to five years,” Marchetti said.

HarbourVest Sees a K-Shaped Software Outcome

John Toomey, CEO of HarbourVest Partners, offered perhaps the clearest articulation of where the private credit community’s thinking has landed on the AI-software question.

He said the software sector is still in the early stages of processing what AI actually means for established business models, noting that the industry realized “almost overnight” that AI could threaten long-standing software businesses. 

However, he argued that sentiment has in some cases overrun fundamentals, leaving investors either too pessimistic about vulnerable companies or insufficiently discriminating about which businesses will genuinely thrive.

Toomey predicted the software sector will mirror the K-shaped economic dynamic that has defined the broader economy throughout the Iran war and post-pandemic recovery, with a divergence between companies that successfully adapt to AI and those that do not.

“Make no mistake, AI is going to affect every business, every industry, every company, and the rate of adoption across each industry and company will increasingly determine their individual success in a competitive world,” Toomey said.

He pushed back against the idea that AI-generated coding tools will allow individual employees to simply build their own software solutions that compete with well-established enterprise platforms, calling that scenario “very far-fetched.”

The emerging consensus across these major private credit managers is that the AI disruption of software is real and accelerating, but that the right response is rigorous underwriting and sector discrimination rather than blanket avoidance of an asset class that has been central to private credit’s growth over the past decade.