AI and SaaS Concentration Risk in Private Credit — Underwriting Deterioration and Maturity Wall

The intersection of private credit and artificial intelligence disruption is producing a structural stress test concentrated in software and Software-as-a-Service (SaaS) lending. While software investments were long considered a highly stable cash cow for direct lenders, the rapid rise of generative AI is disrupting traditional business models, slowing growth, and driving a notable retreat by private credit managers from new software credit estimates.

The AI Threat to Software Collateral

For years, private credit funds aggressively backed software rollups and SaaS companies, attracted by recurring subscription revenues. Today, software represents a massive concentration of collateral, accounting for 19% of total assets for private credit collateralized loan obligations (middle-market CLOs), according to S&P Global Ratings.

However, the rapid advancement of artificial intelligence models is eroding the pricing power of customer-facing software applications. Bank of America analyst Tal Liani highlighted this threat, noting:

"There is increasing risk that AI labs like OpenAI and Anthropic expand downstream into enterprise applications such as Salesforce Inc."

This structural shift is leading to slowing growth, contract cancellations, and margin compression across the software sector. UBS strategist Matthew Mish warned on May 28, 2026, that this AI-driven deterioration will be a primary catalyst for a spike in private credit defaults:

"Our updated perspective points to a meaningful increase in private credit defaults, rising from roughly 4.4% to 9–10%, driven in part by the implications of the AI cycle... Risk is expected to evolve over the next year, intensifying toward year-end and into early/mid-2027 as software businesses experience slowing growth, waning pricing power, margin compression, and contract cancellations."

S&P Data: Managers Retreat from New Software Underwriting

In response to these mounting concentration and AI risks, private credit managers are actively reining in their exposure to new software lending. S&P Global Ratings' SF Credit Brief, published on May 27, 2026, revealed a dramatic pullback in software underwriting:

• The proportion of software issuers among credit-estimated companies fell to 11.2% in April 2026, down sharply from 16.7% in March 2026.

This pullback reflects growing caution among managers who are looking to diversify their portfolios away from software concentration. S&P also noted that software and healthcare issuers reviewed in March and April 2026 had significantly smaller EBITDA sizes compared with previous months, indicating that larger software deals are drying up or returning to traditional syndicated bank markets.

Bank Exposure to Software Debt

The software concentration risk extends directly into the traditional banking sector, which has provided substantial leverage to these funds. Wells Fargo, for example, disclosed in May 2026 that 17% of its $36 billion corporate debt portfolio carries software sector exposure.

As software businesses face slowing growth and contract cancellations, both banks and private credit funds are bracing for a wave of downgrades and restructurings. In its May 26, 2026 financial stability review, the European Central Bank (ECB) simulated a severe shock to the private credit market and specifically warned that "loans to software firms in correlated leveraged debt markets" would be a major transmission channel for contagion, leading to material second-round revaluation losses for insurers and pension funds.