AI Threat Spurs Credit Scrutiny for Software-Focused BDCs, Including Golub Capital BDC

AI tools force fresh scrutiny of software-heavy BDC loan books

Golub Capital BDC and other business development companies face renewed credit scrutiny as new AI tools threaten the business models of enterprise software borrowers that make up a material slice of private credit portfolios. The launch of advanced models by Anthropic is accelerating concerns that software vendors — long a favored borrower group for private lenders — may see revenue erosion if customers adopt AI-driven alternatives that undercut legacy software functions. That raises the prospect of weaker cash flows, covenant pressure and a higher incidence of loan amendments or defaults in the private credit market.

The risk is concentrated because enterprise software accounts for roughly 17% of U.S. BDC investments by deal count, second only to commercial services, and many of those financings take the form of unitranche and other illiquid structures backing leveraged buyouts. Market participants and data provider PitchBook note that unitranche loans have been a hallmark of private credit since 2020, leaving lenders exposed to valuation and liquidity uncertainties if borrower cash generation deteriorates. UBS frames an aggressive disruption scenario in which U.S. private credit default rates rise materially — to about 13% — compared with roughly 8% for leveraged loans and 4% for high-yield bonds, a stress that would notably affect BDCs that concentrate in software.

For Golub Capital BDC, a prominent middle‑market direct lender and business development company, the development prompts active portfolio management and heightened due diligence. Golub is evaluating covenant protections, stress-testing borrower cash flows against accelerated AI adoption, and reassessing pricing and term structures on new deals. At the same time, lenders such as Golub may find opportunities to help viable software firms adapt — through refinancing, add‑on working capital or financing for AI transition projects — while tightening underwriting on companies with limited runway or weak competitive moats.

Concentration and structure amplify the challenge

Analysts and academics caution that private credit portfolios were already showing signs of stress such as liquidity shortfalls and loan extensions before the latest AI concerns, meaning software disruption could act as an additional shock. Questions about loan covenants, secondary market liquidity for unitranche exposures, and the pace at which borrowers can reprice or restructure obligations are now front‑of‑mind for BDC managers.

Broader sector outlook

With roughly $3 trillion in private credit outstanding, the industry is monitoring adoption curves for advanced AI models and their real‑world impact on incumbent software vendors. Lenders and regulators are watching for higher rates of covenant breaches and for deal teams to incorporate AI risk into credit analyses and portfolio allocations going forward.