For years, software lenders relied on recurring revenue, high retention and strong margins as evidence of credit quality. Artificial intelligence is complicating that assumption.
A software-as-a-service (SaaS) borrower can still report healthy annual recurring revenue (ARR) growth and net revenue retention (NRR) even as platform bundling, reductions in paid seats and AI-native competitors begin to erode its long-term competitive position.
The underwriting question is therefore changing. Lenders must assess not only whether a borrower is experiencing financial pressure today but also whether its product, pricing model and customer relationships are becoming structurally less defensible. That matters because private credit has substantial exposure to the software sector. According to a March 2026 analysis by the Bank for International Settlements, outstanding private-credit loans to SaaS companies exceeded $500 billion, or about 19% of direct loans, by the end of 2025.[1]
For lenders, the challenge is no longer simply identifying financial stress but determining whether competitive erosion is taking hold before it appears in traditional credit metrics.
A two-lens view of SaaS credit risk
To explore that risk, we analyzed 133 publicly traded enterprise software companies through two lenses.
The first evaluates observable financial performance across five areas:
- Revenue growth
- Customer retention
- Profitability
- Cash generation
- Leverage
The second assesses AI disruption risk across four areas:
- Product substitution and response
- Revenue-model resilience
- AI delivery capability
- Competitive defensibility
Each lens was scored on a scale of 1 to 4, with higher scores representing greater risk.
The resulting matrix groups companies into four risk profiles:
- Lower relative risk: Stronger financial performance and lower assessed AI risk
- Financially exposed/lower AI risk: Weaker financials, but challenges appear more operational, cyclical or balance-sheet-related than structurally driven
- Latent AI risk: Comparatively healthy financials, but elevated exposure to future pricing, retention, seat or margin pressure
- Dual risk: Financial stress and AI exposure coincide, increasing the risk of reinforcing pressure on growth, margins, liquidity and refinancing
The analysis reveals a visible relationship between higher AI disruption risk and financial stress, but the dispersion is more important for lenders.
Some companies show weak financials despite comparatively low AI risk, suggesting more conventional operating or balance-sheet challenges. Others remain financially healthy while already exhibiting product, pricing or competitive vulnerabilities that may not yet be visible in ARR, NRR or cash flow.
Where AI Risk Appears Most Concentrated
The analysis suggests that AI risk is not distributed uniformly across the software sector. Security, vertical-market software and infrastructure software appear comparatively resilient, supported by mission-critical workflows, specialized data, integration depth and switching costs.
Cybersecurity illustrates why. A competing feature bundled into a larger platform may not be sufficient to displace an incumbent when customers also require proven security performance, regulatory compliance, deep integrations and confidence in the vendor’s ability to respond to emerging threats. Vertical-market software can benefit from similar protections when it serves as a system of record embedded deeply in regulated or specialized workflows.
Enterprise software, enterprise resource planning (ERP), human capital management (HCM) and software-enabled services occupy a more balanced middle ground. Their exposure depends less on the category itself than on whether the product controls a critical workflow, possesses differentiated data and can adapt its pricing and operating model.
Customer relationship management (CRM), analytics, productivity software and software-and-service platforms warrant greater scrutiny. These categories are more exposed to feature commoditization, seat compression, platform bundling and AI-native alternatives.
The implication is that broad software classifications are insufficient for credit analysis. Two borrowers in the same category may have very different long-term credit profiles depending on product substitutability, workflow criticality, data advantages and management’s ability to respond.
These findings also should be treated as directional when applied to private borrowers. Public companies provide more consistent disclosure but may have broader product portfolios, greater access to capital and more resources to invest in AI. The analysis therefore is most useful for identifying underwriting questions and relative risk factors rather than establishing definitive sector rankings.
The Risk May Not Yet Be Visible in the Numbers
Consider an illustrative project-management SaaS borrower growing ARR at 25% with 115% NRR. Traditional metrics suggest a healthy credit profile. But if AI assistants embedded within larger platforms begin replicating core functionality and reducing paid-seat requirements, long-term enterprise value may begin to erode before revenue growth or retention visibly weakens.
By contrast, a vertical-market software company supporting regulated health care workflows might grow more slowly yet remain considerably harder to displace. Its system-of-record position, proprietary data, regulatory complexity and embedded customer workflows could provide stronger long-term defensibility, making the slower-growing company the potentially more durable credit.
From Enterprise Value to Recovery Value
AI disruption can alter a software lender’s downside case even before liquidity becomes acute. Because many software borrowers have limited hard-asset collateral, recovery value often depends heavily on recurring-revenue quality, customer retention, intellectual property and strategic-buyer interest.
If AI-enabled competitors or larger platforms weaken renewal rates, pricing power or paid-seat requirements, both cash flow and exit valuations may come under pressure. Refinancing capacity can deteriorate at the same time that potential recovery value declines.
Lenders should therefore stress downside cases not only for liquidity and covenant headroom but also for lower retention, pricing pressure, incremental AI investment and weaker valuation support. A borrower may remain within covenant limits even as the competitive foundations supporting its enterprise value begin to erode.
What Lenders Should Do Differently
The practical implication is that AI risk should be incorporated throughout the life of a software credit, not assessed only when financial performance begins to deteriorate.
During Origination
Lenders should test whether AI can replicate the borrower’s core customer outcome, whether revenue depends heavily on paid seats, whether larger platforms can bundle competing functionality, whether the company controls differentiated data and whether management has a credible product and monetization roadmap.
During Portfolio Monitoring
ARR and NRR should be supplemented with leading indicators such as seat contraction, discounting, competitive win-loss trends, AI product adoption, AI monetization, gross-margin pressure from inference costs and changes in implementation or support economics. These indicators may reveal weakening competitive position before it appears in headline financial metrics.
During Amendment Discussions
The source of underperformance matters. For financially stressed companies with lower AI risk, conventional actions around liquidity, cost structure, pricing and sales productivity may be more appropriate than fundamental product transformation. For latent- or dual-risk borrowers, financial relief without a credible product, pricing and operating response may simply postpone the underlying problem.
During Refinancing
Lenders should consider whether current growth, retention, margins and valuation assumptions remain sustainable as AI reshapes the category. A borrower that appears financeable on trailing metrics may warrant more conservative assumptions around forward growth, retention, pricing power, valuation multiples and recovery value if competitive defensibility is weakening.
Five AI questions Every Software Lender Should Ask
- Can AI or an AI-enabled platform replicate the borrower’s core customer outcome?
- Is the revenue model vulnerable to seat compression, bundling or changing pricing metrics?
- Does the company control proprietary data, a system of record or a mission-critical workflow?
- Is AI improving customer value and unit economics, or primarily adding cost?
- Does management have a credible product, pricing and capability roadmap?
Outlook
AI will not make every software borrower riskier. It is more likely to increase the dispersion between companies that own critical workflows, data and customer relationships and those whose functionality can increasingly be replicated or bundled elsewhere.
Traditional SaaS metrics remain essential, but they primarily reflect observed operating performance and may lag changes in a borrower’s competitive position. Healthy ARR growth or NRR today does not necessarily indicate that pricing power, product differentiation or customer dependence will remain equally durable.
The AI-risk assessment should therefore be viewed as a screening tool, not the endpoint. For software lenders, the objective is to identify the specific pathways through which AI could affect customer behavior, pricing, margins, refinancing capacity and recovery value, and to monitor those pathways before deterioration becomes visible in headline financial metrics.
Over the next several years, assessing competitive resilience may become as important to software underwriting as ARR growth, NRR and leverage are today. Lenders that identify structural deterioration early will be better positioned to protect enterprise value and credit recoveries.
Justin Eisenband is a senior managing director in FTI Consulting’s Telecom, Media & Technology practice and co-leads the Technology vertical. He has more than 15 years of experience advising companies and capital partners on financial, strategic and operational issues.
Darin Miller is a senior managing director in FTI Consulting’s Telecom, Media & Technology practice and co-leads the Technology vertical. He has more than 25 years of experience in business transformation, transactions and executive leadership.
Andrew Bidylo is a managing director in FTI Consulting’s Telecom, Media & Technology practice. He has more than 15 years of consulting and industry experience advising private equity funds, portfolio companies and corporate clients on operational improvements, growth strategy and M&A execution.
Fabian Namgalies is a managing director in FTI Consulting’s Technology practice, specializing in value creation for private equity-backed and public technology companies. He has more than 20 years of experience across consulting, corporate operations and technology leadership.
The views expressed herein are those of the author(s) and not necessarily the views of FTI Consulting, Inc., its management, its subsidiaries, its affiliates, or its other professionals. FTI Consulting, Inc., including its subsidiaries and affiliates, is a consulting firm and is not a certified public accounting firm or a law firm. FTI Consulting is an independent global business advisory firm dedicated to helping organizations manage change, mitigate risk and resolve disputes: financial, legal, operational, political & regulatory, reputational and transactional. FTI Consulting professionals, located in all major business centers throughout the world, work closely with clients to anticipate, illuminate and overcome complex business challenges and opportunities. ©2026 FTI Consulting, Inc. All rights reserved. fticonsulting.com
[1] Sebastian Doerr, Egemen Eren, Ingomar Krohn and Karamfil Todorov, “Private credit’s software lending meets AI disruption,” BIS Quarterly Review (March 16, 2026), https://www.bis.org/publ/qtrpdf/r_qt2603v.htm .