AI Erodes Software Revenues, Pressures BDC Loans — Hercules Capital Reassesses Underwriting

AI tools put BDCs’ software loans under fresh scrutiny

Anthropic’s rollout of new AI capabilities is prompting a rethink across private credit markets as models begin to encroach on functions that many enterprise software vendors sell, a development that directly affects business development companies (BDCs) and other lenders that concentrate on technology borrowers. Software accounts for roughly 17% of U.S. BDC investments by deal count, second only to commercial services, and industry participants say the potential for AI to compress traditional software revenue streams is raising questions about borrower cash flow durability and loan structures that are already relatively illiquid.

The concern centers on how quickly advanced models can supplant paid software features and professional workflows, reducing recurring revenue and increasing the prospect of covenant breaches or loan extensions for sponsored deals. Private credit is now roughly a $3 trillion market, with many large unitranche loans backing tech buyouts and growth financings. Analysts warn that if adoption of generative AI outpaces borrowers’ ability to adapt, valuation assumptions and covenant protections underpinning those loans may prove inadequate, heightening stress in portfolios that are concentrated in software assets.

For Hercules Capital, a leading BDC that focuses on venture growth and technology companies, the shift crystallizes into a portfolio management challenge. The firm is known for lending to later-stage software and tech-enabled companies that often carry unitranche or subordinated structures; under the current dynamic, Hercules and peers must reassess underwriting assumptions, monitor product displacement risk more closely, and potentially seek stronger covenants or pricing to compensate for elevated operational risk. Market watchers say firms with concentrated exposures face tougher tests of liquidity and workout capabilities as private credit transitions from a growth phase into one where credit performance is more scrutinized.

Market reaction is already evident in wider asset-manager sentiment, as investors and lenders re-evaluate concentration risks in enterprise software. Observers note that strains such as earlier liquidity shortfalls and loan extensions predate the latest AI headline cycle, meaning the sector may be coping with an added shock to already fragile dynamics.

Outlook for lending is tilting toward caution: banks, BDCs and direct lenders are likely to tighten documentation and increase monitoring of enterprise software borrowers while sponsors and management teams rush to demonstrate product differentiation and resilient recurring revenues in the face of accelerating AI adoption.