How has the mortgage industry used chatbots like ChatGPT previously, and what additional AI tools can we expect to see in the future?
AI has been a fashionable topic for a long time now, and not just for technologists. The promises these new technologies make for a better future are compelling. Recently, we’ve seen an explosion of examples of exactly what this future might look like.
Just hop on to any AI blog and you’ll see that new tools are being rolled out on a daily basis, allowing just about anyone to write, draw or even code anything they can imagine.
Naturally, we’re paying close attention to see how these new tools can benefit users of our own Mortgage Cadence Platform, MCP. We have been for some time.
While this may surprise many who are used to thinking that financial services companies lag behind other industries when it comes to technology, we were making good use of chatbots long before ChatGPT came along.
Examples are all over the field, including Bank of America's Erica, Capital One's Eno, Mastercard's KAI and Amex's AskAmex.
Mortgage lenders have been making use of these tools for years, allowing their websites to guide customers into the process and right up to the mortgage application.
Examples in our industry include Ally Home's chatbot, "Ally Assist," the "Rocket Mortgage Assistant," and Mr. Cooper's chatbot, "Cooper."
These tools can provide prospective borrowers with personalized mortgage information and advice. They can answer questions about the mortgage process, provide updates on the status of an application, and connect customers with a loan officer. All without the assistance and involvement of a human.
As these tools continue to advance, we expect to see them used more often by mortgage lenders for customer support and assistance, account management and transactions, fraud prevention and detection and even providing personalized financial advice.
There are plenty of reasons that lenders will embrace this kind of non-human support. By providing consumer support and answers at any time of the day or night, they will see higher customer satisfaction scores. Since they won’t be employing humans in this work, they’ll see better efficiency overall and cost savings.
But there is still a lot of work to be done before the industry can put that much trust in these new tools.
Security and privacy concerns top that list, but there are others, including the work of integrating the tools with legacy systems, creating new processes to allow for human oversight and intervention, and programming the new tools to overcome language and cultural barriers.
But the upside potential is huge. Consider a well-trained and managed chatbot that could guide refinance borrowers through a streamlined process. If the lender already holds the original loan, all the information about the home is already in the platform. All the borrower needs is guidance to progress through the application process.
In the not-too-distant future, the chatbot might work directly with the underwriting department to take the new refinance deal all the way to the closing table.
We’re not there yet, but you can bet that when the tools are ready, we’ll be working on adding their power to our evolving MCP LOS. That’s a topic we look forward to getting into with our lender partners.
By Joe Camerieri, EVP, Sales & Strategy at Mortgage Cadence
Follow us on LinkedIn to be notified when our next article is released.