AI disruption threatens private credit software exposures
A Bank for International Settlements (BIS) Bulletin finds that private credit firms, particularly Business Development Companies (BDCs), have significant exposure to software firms vulnerable to generative AI disruption. While current metrics show no stress, narrowing credit spreads and concentrated lending could amplify future losses.
Software's $115 billion AI challenge
Direct lenders, including private credit firms, have accumulated substantial exposures to the software sector, now facing risks from generative artificial intelligence (AI) disruption.
Business Development Companies (BDCs), which originate a fifth of all US direct loans, have lent approximately $115 billion to software firms.
This represents about a fifth of their total lending and over 80% of their technology portfolios.
Software, once attractive due to recurring subscription revenues, high margins, and low capital needs, is now highly vulnerable to AI.
AI can substitute existing products, lower development costs, reduce barriers to entry, and intensify competition, raising questions about risk pricing and future portfolio performance.
Undifferentiated risk, hidden vulnerabilities
Despite the growing risk of AI disruption, lenders have not charged higher spreads to software borrowers; credit spreads have fallen and converged across sectors by late 2025.
Publicly listed BDCs' valuations also do not reflect their software exposure.
While current performance indicators show no stress, these backward-looking metrics may understate underlying vulnerabilities.
Narrowing spreads reduce loss absorption buffers, and concentrated lending to shared borrowers could lead to simultaneous shocks across many BDCs.
Opaque exposures beyond BDCs could further amplify AI disruption's impact.
A calm before the storm?
This BIS Bulletin highlights a significant, yet currently unpriced, risk building within the private credit market.
The apparent stability of current credit metrics may offer a false sense of security, given the rapid pace of AI innovation.
Regulators should heed these early warnings and consider enhanced transparency requirements to prevent potential systemic issues.