The exit queue at the largest private credit funds has grown measurably since the final quarter of 2025. Redemption requests have climbed quarter over quarter, and several major fund managers have responded by capping quarterly withdrawals—a step that, in this asset class, tends to amplify the anxiety it is meant to contain.
The Trigger: AI Displacement Risk in Software Portfolios
The proximate cause, as analysts covering the space have framed it, is concentrated exposure to enterprise software companies whose revenue depends on adoption patterns that generative AI is beginning to disrupt. Private equity firms loaded their private credit vehicles with loans to mid-market application software companies between 2022 and 2024, many at high leverage multiples. The question circulating through LP committees is whether those borrowers can service that debt if AI code-generation tools eat into their customer bases over the next three years.
Eileen Appelbaum, co-director of the Center for Economic and Policy Research, laid out the structural architecture of the problem in April 2026. Over the prior seven years, private equity firms moved aggressively into the life-insurance and annuity business, redirecting policyholder reserves into private credit funds characterized by thin disclosure and limited mark-to-market discipline. Those same funds extended credit at scale to portfolio companies whose revenue models rest on enterprise software adoption.
Opacity Is the Core Problem
What makes the situation difficult to price from the outside is the disclosure structure of the funds themselves. Private credit vehicles do not report at the loan level. Aggregate exposure to “software” appears as a single line item in most quarterly letters. The breakdown by AI-displacement risk category does not exist in any disclosed format. Limited partners are withdrawing not because they can document losses, but because they cannot estimate the tail.
That uncertainty compounds mechanically. When a fund moves to gate, the gate announcement becomes the news event. Secondary-market buyers price additional discount into fund interests. The cohort of LPs who were watching moves to act. Two of the largest perpetual private credit vehicles disclosed quarterly outflow caps in March 2026. A third followed in early April. None of them have disclosed material credit losses. Investors are pricing the possibility, not a confirmed outcome.
Which Portfolios Carry the Most Exposure
The risk is not uniformly distributed. Portfolios with the highest exposure are those that lent heavily to horizontal application software companies—the kind whose value proposition competes directly with what large language models can now deliver at marginal cost. Portfolios with lower exposure stayed in infrastructure software, vertical SaaS with deep workflow integration, and physical-asset-backed lending where AI substitution is not a near-term factor.
The structural debate among fund managers echoes arguments made about public high-yield credit in 2008 and 2015. The optimistic case holds that direct lending structures, tighter covenants, and private workout processes give managers tools that public bond markets lack. The skeptical case holds that marks have not been stress-tested by a forced-sale environment, and that the workout tools are only as good as the underlying borrower’s ability to generate cash.
What the Next Two Quarters Will Show
The clearest signals will come from NAV prints at the largest perpetual vehicles over the second and third quarters of 2026. Secondary to that: whether LP letters begin carrying explicit AI-displacement-risk metrics. That kind of disclosure does not emerge voluntarily—it appears when the LP base starts demanding it with enough force to move managers to act.
Until then, the redemption cycle has its own momentum. Gates trigger discounts; discounts trigger more redemptions. The asset class is learning, in real time, what it means to have built a transparency deficit into the foundations of its investor base.
Source: Private Credit Fund Redemptions Climb Sharply, Some Caps Now in Place
