AI adoption excites and worries mortgage professionals

While everyone in mortgage has heard of artificial intelligence, and some are actively using it, uncertainty among the industry’s professionals may currently outweigh enthusiasm, remarks in a recent survey suggest.

Still, although doubts about AI may not ever be fully removed, anonymous responses in a survey conducted by Arizent, the parent company of National Mortgage News, about the power of artificial intelligence show widespread recognition of the potential the technology has to transform how the industry works, even when details are not fully understood.    

“I am fascinated by AI and its capabilities. I see it as an excellent tool for automating manual tasks, enhancing efficiencies, identifying trends and issues and raising the level of the entire operation,” one respondent said when describing their level of understanding about artificial intelligence.  

At the same time, they added, “Of course, this can only happen once we resolve the substantial data privacy concerns.”

The buzz surrounding AI, particularly generative and machine learning models, has already led to implementation into some internal tools mortgage employees are using at companies including Blend and Rocket Mortgage. Last fall, both companies introduced AI capabilities aimed at streamlining workflow and answering common questions loan officers posed.  

Artificial intelligence will likely continue making inroads in 2024 for internal functions, typically for educating staff about laws or regulations or assisting them in their interactions with customers. 

“AI is being used in our call centers to transcribe calls, provide feedback on reasons for calls, recognize the customer’s tone and provide solutions to the customer service representative in real time,” one commenter noted.

“AI has the potential to change our landscape, from following up on preapprovals to tracking clients getting ready for a move-up or downsize purchase,” added another.

Generative AI has also proven to be effective among some lenders who utilize it as often as daily in composing content and marketing material, helping with search-engine optimization or producing email responses — but with caveats. 

“AI can be used to help research and write copy for marketing, but the resulting copy will still need to be verified and massaged,” said one user. 

A more cautious tone appears when it comes to unleashing artificial intelligence in any sort of  customer-facing capacity, although a few companies are examining how a tool could be designed to be compliant while still providing value and protection to a client. But the lack of clear regulations did not deter some lenders from testing the waters with AI chatbots in the past year. Others are still “working out requirements to assess where they can potentially be used.”

Some degree of pessimism and skepticism remains prevalent throughout the industry, with job reductions cited as a concern among over half of the mortgage industry professionals taking part in the survey. But across all financial services industries, the loss of personal communication with clients ranked as the leading cause of worry, even as mortgage technology advocates note that certain industry jobs will always require experienced professionals in the process.

Many also did not hesitate to voice what they saw as other possible dangers of widespread AI adoption, such as the loss of skills and knowledge, described by one respondent as the “dumbing down of brain function” with workers “hitting the easy button.” 

“People can use it to mask their incompetence,” another added. 

And even while nobody in the mortgage industry has advocated for the use of AI in approving or denying loan applications, risk may lie in its potential to influence the decision making process over the long term, some fear.

According to one mortgage professional, a future could emerge where lending turns into too much of “a black-and-white process that can inadvertently cause a decline in the acceptance of loans outside the credit box or that have certain characteristics that need a more nuanced approach. This then becomes a self fulfilling prophecy that further digs the lender/broker into a small box.”

Another said they thought generative AI had a tendency to reply with the most popular answer to a question, but “this often is not the best or right answer to a question.”  

But even mortgage professionals who raise concerns about the rise of AI acknowledge its “implications are tremendous,” with some expecting it to be “a huge part of the business in 2024.” 

“It will definitely help in getting the mortgage processing timelines reduced and make the process of taking a mortgage more easier for consumers,” a respondent said.