Real Estate
Home›Real Estate›Industry›Mortgage lenders increasingly use AI for document inde…
Mortgage lenders increasingly use AI for document indexing and underwriting
A new STRATMOR survey finds 68% of lenders use AI to classify and index documents, while integration gaps can force manual reconciliation later in the loan process.
Mortgage lenders are expanding the use of AI in parts of the loan origination workflow, but technology disconnects can limit the payoff, according to HousingWire, citing STRATMOR Group’s 2025 Technology Insight Study. The study highlights progress on document handling, with 68% of lenders using AI to classify and index documents, 59% using it to read documents, and nearly 50% using it to analyze borrower income during underwriting.
HousingWire reports that the value of those upstream AI steps depends on whether the information can flow through the rest of the origination lifecycle. If AI-generated data cannot move seamlessly into downstream systems, lenders may still need to reconcile details manually during underwriting, even when a document is correctly classified earlier.
The outlet also points to a common systems issue within lender technology stacks, where point of sale, origination, underwriting and document management systems do not always exchange data smoothly. That disconnect can carry discrepancies into closing, potentially leading to problems that only surface during post-close quality control or investor review.
HousingWire notes that regulatory requirements such as TRID, RESPA, HMDA and fair lending rules raise the stakes for accurate, traceable documentation, and that manual checks may continue because staff still must verify where systems stop talking to each other. As a result, manual review time can scale with loan volume rather than fully shrinking as AI adoption increases.