Private credit fund managers built the asset class’s reputation on controlled environments: private borrower relationships, negotiated covenants, marks set without the noise of daily market fluctuation. That controlled environment is now generating a problem it was not designed to address—a redemption wave driven by a risk category (AI-displacement in software) that the controlled disclosure structure cannot quantify for the LP base that is trying to price it.

Three Gates in March and April

Two of the largest perpetual private credit vehicles disclosed quarterly withdrawal caps in March 2026. A third followed in early April. The combined timeline—three gates in approximately six weeks—is unusual enough in the asset class to constitute its own news event, separate from the underlying credit question. The funds communicated the caps as standard liquidity management. Secondary market participants communicated through pricing: discounts on fund interests widened with each announcement, reflecting the market’s independent assessment that stated NAVs may not fully capture where marks will eventually land.

None of the three gated funds have disclosed material credit losses. The gates are responding to redemption queue volumes, not to confirmed borrower defaults. The LP base filing those requests is exiting because it cannot price the AI-displacement risk from available fund disclosures—not because it has seen the credit losses it is worried about.

The Insurance-Capital Chain

The background to the exposure is the PE-insurance-private credit structure that Eileen Appelbaum of the Center for Economic and Policy Research documented in April 2026. PE firms acquiring life-insurance and annuity businesses over the prior seven years gained access to policyholder reserves—long-duration, stable capital. Those reserves funded proprietary private credit funds with minimal disclosure obligations and infrequent marks. The credit funds deployed into PE-owned portfolio companies, with the 2022–2024 vintage concentrated in mid-market software businesses that took on high leverage based on subscription revenue assumptions that the current AI environment is testing.

The structural concern the CEPR analysis raises is not just investment risk—it is a question about disclosure adequacy across the chain. Policyholders whose reserves back these portfolios have no direct line to the credit quality of the underlying software loans. The LP investors one level up have fund-letter disclosure that aggregates software exposure without sub-category detail. Neither layer can measure the AI-displacement risk from the available information.

The Portfolio Differentiation That Gets Lost in Aggregate Stats

Aggregate private credit statistics on software exposure obscure the risk differentiation that matters for credit analysis. Portfolios that lent to horizontal application software—the category where AI substitution is most direct—in the 2022–2024 period carry substantially different risk profiles than portfolios that concentrated in infrastructure software or asset-backed lending outside of technology. That difference is not visible from standard fund disclosure.

The LP community trying to assess its exposure is therefore doing primary research that the fund structure should be providing: reaching out to GPs directly, demanding sub-category breakdowns that are not in standard LP letters, and in many cases concluding that the uncertainty justifies filing a redemption request rather than waiting for better disclosure.

What the Next Two Quarters Will Show

NAV prints from the gated vehicles will provide the first hard signal. If marks hold at current levels through Q2 and Q3 2026, the secondary market will have been overcorrecting. If marks move, the LP base that gated was right to exit at stated NAV rather than accept secondary discounts.

The disclosure question is slower-moving. AI-displacement-risk metrics by portfolio sub-category do not appear in any major private credit LP letter as of April 2026. When they begin appearing—and the evidence suggests LP pressure is building to a level that will eventually force a change in reporting practice—it will mark a structural upgrade in the information available to the market. It will also arrive after the fact, as disclosure upgrades historically do in private markets.

The managers who get ahead of that upgrade—who start reporting AI-exposure sub-categories before LPs force them to—will have a meaningful competitive advantage in the next fundraise cycle. The ones who wait will face the question from their LP base in the context of whatever the marks show by then.

Source: Private Credit Fund Redemptions Climb Sharply, Some Caps Now in Place