In recent years AI has been seen as a potential catalyst for software companies—a way to reinvigorate growth rates while also expanding margins. But public markets have recently begun to question that thesis: Even where competitive moats exist, can software companies realistically return to stronger growth? And if not, how should investors think about the terminal value of those companies?
Recent product releases from OpenAI and Anthropic have sharpened these concerns and continue to rattle broader investor confidence in the sector. The questions are legitimate: Is AI weakening the moats that made software so attractive in the first place? Are private valuations next to fall? And what does this mean for software equity investors and credit investors and lenders?
Not All Software Is Created Equal
AI is unquestionably disruptive, but disruption is rarely uniform. While we believe software companies that provide single-function, point solutions may be vulnerable to replacement by a capable AI model, we feel those that offer embedded, mission-critical platforms—systems with deep workflow integration, strong retention and proprietary data advantages—may prove more resilient.
This distinction matters for investors: In our view, a portfolio built on differentiated, high-retention software businesses with genuine data moats could differ meaningfully from one weighted toward generic tools, and should be evaluated accordingly.
Private Markets Are Not the Public Market
Public market volatility is real: It directly influences sentiment and public comparables used in quarterly valuation frameworks, particularly when it continues over longer periods, and may well affect future quarterly valuations of private equity-owned software companies.
However, we believe public market volatility is not necessarily a direct read-through to private valuations, and that company fundamentals—including growth rates, margins, depth of customer relationships and long-term positioning—will in the long run also be critical in determining company valuations. Indeed, we find that private equity ownership can itself be a meaningful buffer because skilled sponsors may be able to invest through volatility, support management teams and reposition businesses in ways that public market participants often cannot.
Although we can’t predict the long-term disruption that AI may cause, we believe that actively managed, privately owned software businesses with a differentiated product, high retention rates, a critical function and/or data moat remain well positioned to grow in the short to medium term.
Financing: Where Discipline Matters Most
Meanwhile, the credit landscape has grown more complex—and that’s when our experience suggests active management can make a meaningful difference.
In our view, businesses where AI strengthens product differentiation and efficiency should continue to access financing on reasonable terms, while areas where AI could accelerate commoditization or compress pricing may see tighter structures, lower leverage and a higher all-in cost of capital.
One area that we feel requires particular attention: 2021 – 2022 vintage loans and associated equity approaching maturity. Where growth has disappointed and leverage has not come down as expected, we believe that refinancing risk has risen. While pay-in-kind (PIK) structures have helped preserve near-term liquidity, we find these can compound leverage over time and reduce future refinancing flexibility if growth assumptions don't materialize. In this environment, rigorous scenario analysis remains paramount.
How to Think About AI in Underwriting
At Neuberger Private Markets, we believe in assessing AI risk and opportunity in two directions for every investment:
- Risk lens: Is AI changing customer workflows in ways that erode a product’s value? Could AI compress pricing, lower barriers to entry or enable a competitor to replicate core functionality?
- Opportunity lens: Is AI expanding this company’s addressable market? Could it improve retention, enhance product capabilities or reinforce the competitive moat?
In our view, this two-directional discipline—applied consistently at the underwriting stage—can help mitigate risk and ultimately shape which businesses are positioned to generate durable returns through this period of disruption.
The Bottom Line
We believe ongoing AI disruption could play to the strengths of experienced active managers able to target privately held software players that appear well-positioned to capitalize on AI rather than be supplanted by it. This includes companies that provide mission-critical systems or those that have data moats and network-based barriers to entry. Despite recent market volatility, we believe this environment may continue to create attractive opportunities for experienced, rigorous underwriters with the conviction to distinguish the durable from the vulnerable.
