Lenders lag equity on AI underwriting

Commercial real estate equity investors have run deals through artificial intelligence for several years now, sifting portfolios and flagging problems before they ever reach an investment committee. Lenders are only starting to catch up, and the reason, according to Laura Krashakova, founder and CEO of Smart Capital Center, an AI-powered underwriting and portfolio-monitoring platform, has less to do with the tools than with who stands closest to the hole.

If a property's value falls 20 percent, Krashakova points out, the equity investor absorbs the loss and, in a bad enough outcome, the whole property; a lender at 50 to 60 percent loan-to-value may lose nothing at all. That uneven downside exposure produced an asymmetry in urgency, which is the single most useful fact in the story: technology adoption in commercial real estate credit has tracked the pain, not the promise.

Compliance has widened the gap, because lending institutions carry regulatory requirements that equity investors do not, and those requirements slow how quickly new tools can be pushed through underwriting and credit-decisioning workflows. That is a real constraint, but it is also a convenient one—a reason to move deliberately that doubles as a reason not to move.

For all the talk about regulation and risk exposure, what is finally forcing the debt side forward is the maturity wall and the fundraising market. Lenders and credit funds are staring down a heavy schedule of loan maturities, tighter conditions for raising capital, and investors who want reporting turned around faster than the current process allows, while private lenders and credit funds have drawn heightened scrutiny over underwriting and credit-decisioning practices through the past two years of distress in private lending.

Put those forces together and the incentive structure inverts: the equity investor's AI spend buys return; the lender's buys defense—documentation of how a loan was sized, how a portfolio is monitored, and how fast a borrower's deterioration shows up in the numbers. As PWD has argued, the refinancing wall is a duration transfer from banks to private credit, and only vehicles that can wait out a maturity will own the next leg; sitting through a maturity is a monitoring problem before it is a capital problem, because a lender that cannot see deterioration in its book in real time is not patiently holding an asset, it is holding an unknown.

The way a credit shop frames the tooling now decides what it buys. If AI underwriting gets budgeted as a reporting and compliance expense, it lands in the same bucket as the disclosures—necessary, unloved, and disconnected from the decision to lend; if it gets wired into credit decisioning itself, the shop gets the thing that actually prices: earlier sight of which borrowers are drifting and the ability to act on it before the extension conversation. The equity side made that call years ago, and the debt side is being told to make it now, with the maturity wall already on the schedule.