AI: A Great Tool (With Limitations)

The impact of AI on the mortgage industry is undeniable and it will be fascinating to see how it continues to evolve the sector in the coming years. That’s why I’ve asked three mortgage industry experts to share their thoughts and insights on the current state and future of AI in the sector. Check out their valuable perspectives on this exciting topic.

Tammi Alexander
Senior Director
Lender Toolkit

AI is truly transforming the mortgage industry and forever changing the homebuying process. AI technology provides a wealth of benefits by quickly processing and analyzing large amounts of data. Automating tasks like verifying documentation, comparing financial factors, and analyzing collateral data not only saves time but also improves the accuracy of information communicated. AI-powered systems can provide informed decisions about creditworthiness, collateral value, and risk assessment, reducing the time to funding, improving loan delivery quality and minimizing risks.

The potential for AI to dramatically enhance the efficiency and accuracy of mortgage lending while reducing costs and minimizing risks is significant. Companies like Lender Toolkit are at the forefront of implementing these innovative AI solutions to improve the customer experience and drive business growth. If companies aren’t embracing AI, they need to.

Faith Schwartz
CEO
HousingFinanceStrategies.com

Data-driven insights will start to streamline the mortgage efforts, and the Federal Housing Finance Agency’s new office of technology should help spur movement. Given some of the leadership through the GSEs, [government sponsored entities] we have nimble tech, source data, and driving front-end solutions that have been traditionally paper intensive. Source data can drive some great results. Think about having access to borrower permissioned data, such as bank statements, tax returns, employment and income status at the time of application. You have most of the important information process and underwrite the loan. Pre- and post-funding data audits can support this approach.

With 3-D scanning, hybrid appraisals and acceptable AVMs [automated valuation models] for collateral, you also have data-driven solutions to help determine collateral risk. The challenge is accessing legacy systems, workflow, and the data, and ensuring all the systems and compliance to the laws are in place. A very exciting time in mortgage that is filled with promise.

Maya Gardner
Director, Program Management
Blue Sage Solutions

With the rising interest rate environment and overall contraction in the market, lenders are looking for any number of technologies and process improvements to provide increases in productivity at lower costs. Just as rules engines and automated underwriting engines have been used in the past to assist with approvals and closing loans faster, AI and machine learning (ML) also have had a significant impact on the industry in a number of ways. Automated document classification and data extraction are both AI-enabled functions that have already transformed the mortgage process, especially in the correspondent purchase and QC review process to eliminate “stare and compare” processes. Chatbots also have provided enhanced customer service capabilities, allowing lenders to provide responses to questions and expertise 24/7.

This year, lenders will be leveraging AI even more and more in underwriting and other tasks, such as fraud detection. Using a vast amount of historical data, sometimes over many years, AI and ML tools can generate algorithms which can provide an automated decision and a series of why the decision was made using a number of supporting factors, reducing the dependency on manual processes. These models are continuously updated as more loans are originated, approved, and closed, becoming faster and smarter over time. Other key aspects of the mortgage process such as managing risk, reducing fraud, and reducing default rates all benefit from AI models.