The big interview: AI making private lending faster is making fraud easier, CEO warns
One CEO in the private lending technology space said the shift is accelerating faster than most lenders recognize, and that building systems capable of catching it before a bad loan closes is now a competitive requirement.
Sourabh Chirimar (pictured top), CEO of The Mortgage Office, said what he is seeing in private lending is not organized fraud but something more opportunistic and harder to catch.
“AI has made it very easy to gloss over things,” Chirimar told Mortgage Professional America. “It’s very easy to generate fraudulent documents. A lot of this market is lending to LLCs, and it’s easy to just create a new LLC, fabricate a track record or an appraisal or anything. There’s just a lot more cases of what I would call a micro level of fraud, where they don’t have the document, so it’s made up with AI.”
Why private lending is targeted
Chirimar said the structural characteristics of the private lending market make it particularly vulnerable. The loans are short, typically six- to nine-month interest-only flip loans, which means due diligence windows are compressed, and borrowers are often newly formed LLCs with limited verifiable history.
He said the goal is to shift lenders from reactive to proactive risk management, flagging borrower distress signals across the broader data set before they show up in the lender’s own portfolio. The fraud detection layer works similarly, allowing lenders with access to cross-portfolio data to identify patterns such as new LLCs with suspiciously similar documentation, appraisals that don’t align with comparable sales data, or borrower histories that don’t hold up against external records.
