What caused private credit lenders to sour on software companies, how large is the exposure, what are the signs of a retreat, and what refinancing risks lie ahead as AI disruption threatens SaaS business models?

Private credit lenders have turned sharply against software companies because generative AI breakthroughs threaten to commoditize or replace the core products of many SaaS firms, making their revenue models — and ability to service debt — much more uncertain. Here is a breakdown of the causes, exposure size, retreat signs, and refinancing risks ahead.

What caused the souring

• AI disruption of the SaaS model. In early 2026, Anthropic released Claude Cowork and OpenAI launched new models that let individual users build programs capable of replacing services provided by many existing software companies. This triggered a global sell-off in publicly traded software stocks and raised existential questions about SaaS subscription economics .
• Collateral markdowns by banks. JPMorgan marked down the value of collateral on software loans in private credit financing facilities, reducing the funding available to private credit lenders and causing observers to question portfolio valuations .
• Spiking defaults and "shadow defaults." Several large private-credit-backed companies defaulted (e.g., First Brands Group with ~$10B in debt), and lenders increasingly used "amend-and-extend" maneuvers to mask distress rather than recognize losses .
• Higher-for-longer interest rates compressed software growth rates and valuations, exposing how much debt had been layered onto companies that could no longer grow into their leverage .

How large is the exposure

Estimates vary by how broadly "software" is defined, but the figures are sizable:

• Outstanding loans to SaaS firms grew from ~$8 billion in 2015 to over $500 billion by end-2025, representing ~19% of total direct loans . About a third of all private credit funds had extended loans to the sector .
• BDC-specific exposure: Business development companies — the most transparent segment of private credit — had lent roughly $115 billion to software firms, about a fifth of all BDC lending . Software names accounted for more than a quarter of all companies under credit pressure in BDC portfolios .
• UBS estimates that 25–35% of private credit portfolios face elevated AI disruption risk .
• The broader global private credit market stood at roughly $2.3–3 trillion in AUM as of 2025–2026 .

Caveat: KBRA notes that pure SaaS borrowers were only ~8% of new private credit loans originated in 2025, suggesting some of the headline "$500 billion+" figure may include software-adjacent or tech-enabled services, not just classic SaaS .

Signs of a retreat

• CLO managers offloading software loans. CLO managers have sold software loans at a discount to reduce exposure . Across leveraged loans, junk bonds, and CLOs, the selloff in software debt accelerated sharply in 2026 .
• Redemption surges and gating. Investor redemption requests from semi-liquid BDC vehicles spiked in late 2025 and continued through early 2026 . Blue Owl restricted redemptions in one of its credit funds; five of six major private credit funds blocked investor withdrawals in Q1 2026 in the largest "gating" wave since 2022 .
• Fund-raising shrinkage. Ares Management was forced to shrink a planned fund, and overall capital flows into private credit slowed .
• Lenders pivoting away from software. PitchBook notes that lenders are now actively shifting capital toward industries viewed as more insulated from rapid technological change .
• Distress markers climbing. By March 2026, 10.6% of companies in BDC portfolios showed credit pressure, with software representing the largest share of stressed names . Lower-middle-market loans showed 13% of loans trading below 90 cents on the dollar — an "alarming" distress signal .

Refinancing risks ahead

• A wall of maturities approaches. The massive buildup in software-directed loans from 2019–2022 is now approaching its maturity wall. Companies that borrowed at peak valuations on the assumption of continued high growth must refinance in a market where AI has made their business models less certain and lenders have pulled back .
• Low recovery expectations. Voya Investment Management warns that even though software loans are labeled "senior secured," they are likely to have very low recovery rates in a disruption scenario, because the primary collateral — recurring revenue streams — may deteriorate rapidly .
• Forecast default rates. Moody's stress scenarios project an 8–13% default rate range for AI-exposed private credit borrowers . The BIS has flagged that losses concentrated in the software sector could cascade through the private credit ecosystem if redemption pressures force asset sales into illiquid markets .
• "Amend-and-extend" masking true distress. Many lenders have extended maturities without marking loans to market, kicking the refinancing problem into future years and creating a "shadow default" overhang .
• Bifurcated market risk. Institutional direct lending is still functioning, but retail-facing semi-liquid BDC vehicles face a structural confidence crisis, making it harder for smaller software borrowers to roll over debt .