Private Credit and Software Risk in the Age of AI

February 2026

Introduction

Software has become one of the largest borrower segments in private credit portfolios, but the risk profile of software lending is changing quickly. Artificial intelligence is altering how software companies compete, price, retain customers, and generate cash flow. While AI creates long- term winners, it also compresses margins, weakens moats, and shortens product cycles for many software businesses. 

For private credit lenders, this matters because much of the underwriting framework for software loans was built on assumptions that no longer fully hold. Recurring revenue was historically treated as durable revenue. Customer retention was assumed to be sticky. Pricing power was often underwritten as stable. AI is challenging each of those assumptions at the same time. 

As software exposure has grown across large private credit platforms, these shifts raise important questions about concentration risk, underwriting discipline, and default behavior. The issue is not whether software remains a viable lending category. It is whether lenders are properly accounting for how quickly software fundamentals can change in an AI-driven environment. 

Why Software Became a Core Private Credit Borrower

Private credit moved into software lending for understandable reasons. Subscription revenue is visible and measurable. Gross margins have historically been high. Software businesses are asset-light, allowing leverage to be sized to revenue durability rather than physical collateral. For lenders seeking scale, software offered a large and growing universe of sponsor-backed borrowers. 

Over time, underwriting evolved toward recurring revenue metrics such as ARR, retention, churn, and cohort behavior. This helped support the expansion of ARR-based lending and recurring revenue facilities, particularly for vertical software platforms and software enabled rollups. 

As private credit scaled, software also became attractive from a portfolio construction perspective. It was viewed as less cyclical than capital intensive industries and less exposed to traditional commodity or inventory risk. That perception helped push software into top industry allocations at many large private credit platforms. 

What has changed is not the importance of software, but the stability of the assumptions behind it. From a due diligence standpoint, we view this as a reason to examine how software exposure is constructed, not just how much of it exists.

Software Exposure at Large Private Credit Platforms

ARES (ARCC)                       ████████████████████████   ~23%

BLACKSTONE (BXSL)      ████████████████████       ~20%

KKR (FSK)                              █████████████████          ~17%

BLUE OWL (OBDC)          ████████████               ~12%

APOLLO (MFIC)                 ███████████                ~11%

Source: company disclosures as of Sept. 30, 2025. 

How AI Is Changing Software Business Risk

Artificial intelligence is fundamentally altering the economics of many software businesses, and not always in ways that benefit lenders. In some cases, AI strengthens companies with proprietary data, deeply embedded workflows, or regulatory complexity. But for a large portion of the software universe, AI introduces real competitive pressure. Agentic and generative tools increasingly replicate workflows that once required dedicated applications. Thin moat point solutions face higher churn risk as functionality is bundled into broader platforms or replaced by AI native alternatives. 

Pricing dynamics are also shifting. AI is accelerating the move away from per seat pricing toward usage or outcome-based models. While this can create upside in strong adoption scenarios, it also introduces greater revenue variability and reduces the predictability lenders historically relied on. Renewal pricing can reset faster than expected as customers reassess software spend in an AI-enabled environment. 

Margins are under pressure as well. AI brings real costs tied to compute, inference, and infrastructure. When those costs are not fully passed through, a software company can report stable or growing revenue while free cash flow deteriorates. For credit investors, that divergence matters more than headline ARR. 

The net effect is that revenue visibility has become less synonymous with cash flow durability. Software companies can look healthy on the surface while underlying economics weaken faster than historical patterns would suggest.

What Recent AI Headlines Reveal About Software Risk

Software risk is changing faster than traditional underwriting frameworks were designed to capture. Recent AI-related headlines provide a real time example of how quickly perception around software durability can shift even in the absence of direct credit events. 

A useful case study comes from product announcements by Anthropic. There is no evidence that software companies were operationally or financially impacted by any specific Anthropic release. There were no reported system failures, security issues, customer losses, or changes to Intuit’s credit profile tied to those announcements. 

Yet the market reaction was immediate. Following recent AI product releases from Anthropic, publicly traded software stocks broadly sold off, including established incumbents such as Intuit, Adobe, and Salesforce. These moves reflected a reassessment of long-term growth, pricing power, and competitive positioning for traditional software business models as AI capabilities accelerate. 

Recent market activity has shown that public equity prices for the major private credit managers with significant software exposure have come under pressure alongside software sector volatility. Stocks of alternative credit managers and business development companies (BDCs) have declined as investors reassess both credit quality and concentration risk tied to software and technology loans. In early February 2026, shares of Blackstone were reported down roughly 5% as concerns about software exposure spread with Apollo and KKR also posting notable declines.  

Blue Owl Capital, a firm with some of the greatest relative exposure to private credit and softwarerelated loans, experienced a sharper sell-off, trading down more than 10% in a session and hitting multiyear lows as investors weighed the implications of software risk. Ares Capital Corporation, one of the largest BDCs and often viewed as more resilient, showed more modest relative weakness but still traded lower amidst the broader sell-off in the sector.  

Importantly, these stock declines were driven by valuation and risk perception, not by defaults, covenant breaches, or operational disruptions. The businesses continued to perform, but investor expectations around durability shifted quickly. 

This distinction matters for private credit. Public markets tend to reprice risk immediately, while private credit marks adjust more slowly. However, these equity moves serve as early signals. They highlight how assumptions that once felt stable, such as recurring revenue stickiness, renewal pricing power, and customer switching costs, are increasingly being questioned in an AI-driven environment. 

What This Means for Private Credit Defaults

Software loans tend to fail differently than traditional cash-flow credits. Defaults often follow rapid changes in customer behavior rather than slow operational decline. Retention can fall quickly after product displacement, pricing resets, security issues, or competitive disruption. Once churn accelerates, leverage that once appeared conservative can become unsustainable in a short period of time. 

AI makes these outcomes more likely by accelerating competitive pressure, compressing product differentiation, and lowering customer switching costs. As a result, downside scenarios can materialize with less warning and less opportunity for incremental intervention. 

This dynamic raises the risk that historical loss assumptions understate future default volatility in software-heavy private credit portfolios, particularly where leverage was set using legacy SaaS benchmarks. 

Implications for Underwriting and Portfolio Construction

The takeaway is not that software lending should be avoided. It should be underwritten differently. More resilient software borrowers tend to have durable data advantages, embedded or mission critical workflows, regulatory complexity, or control over distribution. These characteristics make displacement more difficult and revenue more defensible. 

More vulnerable borrowers are often narrow point solutions where functionality can be replicated quickly or bundled elsewhere. In those cases, historical retention metrics may overstate future durability, and leverage needs to be set accordingly. 

For private credit lenders, this environment demands tighter monitoring and more conservative leverage tied to cash flow conversion rather than revenue alone and a willingness to reassess exposure as competitive dynamics evolve. For allocators, it reinforces the importance of manager selection and specialization rather than relying on scale or headline diversification. 

How Divergent Approaches Due Diligence to Minimize These Risks

At Divergent Capital Asset Management, we don’t think about private credit manager selection as a branding exercise or a race to gather assets. We simply treat it as a risk management decision. The goal of our diligence process is to understand risk before it shows up in performance, not after.

We always start with exposure, not yield. Before we look at returns, we focus on what a manager actually owns. That means digging into sector concentrations, software exposure, and overlap with other large platforms. When software is a meaningful part of the portfolio, we go a layer deeper and look at the types of software businesses being financed not just how much capital is allocated to the sector.

From there, we spend a lot of time on underwriting discipline. We want to understand how managers think about software risk in real terms: how leverage is sized, how retention and churn are analyzed, and how assumptions change as business models evolve. Managers who rely on static SaaS frameworks or historical averages tend to struggle when conditions shift. We prefer teams that adjust structures, covenants, and leverage as fundamentals change rather than forcing yesterday’s assumptions onto today’s risks. 

We also pay close attention to how managers behave when things don’t go according to plan. Downside matters more than base cases. We look at how quickly managers identify deterioration, how they respond to early warning signs like pricing concessions or declining retention, and how they’ve handled restructurings and non-accruals in the past. In software credit, early intervention often makes a bigger difference than headline recovery rates.

Finally, we care a lot about specialization, alignment, and liquidity. As private credit has scaled, underwriting frameworks have started to look more alike, with many large platforms financing the same sponsors, sectors, and business models. We intentionally seek out specialist managers with deep domain expertise and long-duration institutional capital.  We focus on teams whose incentives are aligned with capital preservation, not asset growth. And we always evaluate liquidity as part of the investment itself, stress testing redemption scenarios and looking at how funds behaved in prior periods of volatility. Liquidity only matters if it holds up when it’s actually needed. 

Why This Matters for RIAs and Their Clients

For RIAs, the risk is not simply underperformance. It is client experience. 

When clients allocate to private credit, they often expect income stability, diversification, and periodic liquidity. When software-driven stress leads to weaker performance and gates, those expectations are challenged abruptly. 

Our goal at Divergent is not to avoid private credit, but to help advisors understand where risk is building and how to manage it proactively before those conversations are forced by events. Conclusion

AI is reshaping software economics faster than many underwriting frameworks were designed to handle. For private credit, that shift increases the risk that revenue-based assumptions overstate durability and understate default risk. 

Software remains an important borrower segment, but the definition of “recurring” has changed. Revenue visibility no longer guarantees cash-flow stability, and competitive moats can compress quickly in an AI-driven market. 

The right response is not to avoid software exposure, but to be more intentional about how it is underwritten, structured, and monitored. In our view, this environment reinforces the value of curation, specialization, and disciplined due diligence particularly as private credit continues to scale.

For RIAs and allocators, the risk is not simply underperformance. It is concentration risk emerging across portfolios at the same time, driven by shared exposure to software business models undergoing rapid change. Understanding where that risk is building and how managers are adjusting to it has become essential to managing private credit allocations responsibly. 

About Divergent Capital Asset Management 

Divergent Capital Asset Management helps RIAs and family offices access and build customized portfolios of private market investments including private credit, private equity, venture capital, and real estate. The firm provides institutional quality infrastructure for sourcing, structuring, and managing alternative investments. They handle everything from due diligence to administration, allowing advisors to offer branded, turnkey private market solutions. Divergent’s platform bridges the gap between institutional access and independent advisor needs.    

divergentcapitalam.com 

Divergent Capital Asset Management LLC (“DCAM”) is a registered investment advisor offering advisory services in the states of Georgia, Florida, Texas, Louisiana, Colorado, California and in other jurisdictions where exempted. Registration does not imply a certain level of skill or training. None of the information provided is intended as investment, tax, accounting or legal advice, as an offer or solicitation of an offer to buy or sell, or as an endorsement of any company, security, fund, or other securities or non-securities offering. A copy of DCAM’s current written disclosure statement discussing DCAM’s business operations, services, and fees is available on the SEC’s website at www.adviserinfo.sec.gov or from DCAM upon written request.