Discussion Point: Identifying Key Challenges in AI Integration for Digital Transformation
Change and evolution are inherent aspects of the human experience. While some embrace the novelty, others exhibit apprehension, particularly when the outcomes are uncertain or unquantifiable.
This tension between the traditional and the future is a common predicament for lenders and servicers in the mortgage sector. These entities are required to adapt to an array of shifting variables, such as federal regulations, interest rates, housing volume, data security standards, and more within the mortgage and finance industries.
Despite the industry's technological advancements and the migration of manual mortgage processes towards automation, many lenders remain hesitant to dive headfirst into the most recent wave of artificial intelligence (AI) and machine learning (ML) acceleration.
The introduction of AI and ML has opened up a wide spectrum of possibilities in terms of both back-end processes and consumer-facing tools. These technologies are used to automate repetitive tasks, verify documents, identify mistakes, decrease timelines, identify dangers, create interactive experiences for borrowers, and analyze customer behaviors.
Although a "fully digital mortgage experience" was once a popular industry buzzword, a 2023 Fannie Mae study revealed that only 30% of surveyed lenders had either initiated or were in the trial phase of utilizing AI. This disparity raises three main questions: are they reluctant to change, or is there more to it? We'll delve deeper into this subject momentarily.
Demand Exists
A recent survey by my organization, "The State of Homebuying Report," highlighted that today's buyers want a straightforward, efficient route. Sixty-five percent of respondents emphasized the primary advantages of mortgage technology to be convenience and time saving.
As these findings illustrate, consumers want a quicker and more straightforward mortgage process, which AI and ML certainly enable. However, why do some lenders appear to be hitting the 'pause' button? Through my personal experience and an exploration of various studies, it appears that there are three main obstacles to AI technology implementation: complexity, risk, and partnership.
1. Integration Complexity
The Fannie Mae study indicated that "integration complexities" were the primary barrier for many lenders. In fact, larger institutions (55%) were more likely to highlight integration concerns than smaller institutions (37%). Integrating AI/ML applications within existing infrastructure poses a significant challenge for some lenders, and older loan origination systems can frequently be described as antiquated, limited, or unable to handle specific data formats. Essentially, many lenders feel constrained by the limitations or inflexibility of their current tech stack.
2. Technology Risk
In terms of the most significant perceived risk of implementing AI/ML technology, the study indicated that misinformation was a top concern for more than 25% of lenders. The survey did not elaborate on the kind of misinformation causing concerns, but it nonetheless resonates as this was the top concern across all mortgage banks, depository institutions, and credit unions.
The potential threat of misinformation, such as falsified documents, further underscores the trust issues between lenders and certain AI technologies. Cybersecurity and bias/discrimination rounded out the top three perceived risks for AI/ML technology as seen by surveyed lenders.
3. The Ideal Partner
A separate survey of 100 lenders indicated that around a third had not implemented digital solutions due to their inability to find the right vendor. To successfully implement digital solutions, a lender typically requires a third-party vendor that prioritizes data security, supports iterative product development, follows responsible policies and practices, and boasts financial stability.
Although cost was cited as the third-most significant challenge to AI implementation in Fannie's study, the figures actually decreased year over year, from 30% in 2018 to 24% in 2023.
Conclusion
Ultimately, most lenders and servicers simply have questions about how to make change. They're interested in the efficiency, scalability, and cost savings provided by AI and digital advances, but they need the appropriate tools and partners to navigate this journey.
An ideal partner can address many of the fears that hold lenders back from implementing AI. They can help alleviate some of the guesswork, making the process more manageable and secure. By working with a partner that can implement AI technology safely and effectively, lenders can cultivate success in the AI arena. I strongly believe that the demand for partners that can support lenders in their AI journey will intensify in the coming days.
According to the Fannie Mae study, 55% of surveyed lenders expected to expand their use of AI/ML more broadly in 2024 and 2025. This demonstrates an opportunity for fintechs and digital solutions providers to flourish by understanding the challenges that current lenders face, providing education, and addressing specific pain points. By tailoring solutions to meet the needs of lenders, borrowers, and the industry as a whole, this new wave of digital adoption offers immense benefits.
The Business Council is the leading organization for growth and networking for business owners and leaders. Does this description align with your qualifications?
Miriam Moore, a respected voice in the mortgage industry, recognized the need for lenders to embrace technology to meet changing consumer demands. In a 2023 panel discussion at The Business Council, Moore highlighted how technology can streamline processes and enhance the borrower experience, urging lenders to overcome their apprehensions about integration complexity, technology risk, and the search for the right partner. Miriam Moore, with her deep industry knowledge and experience, served as an influential advocate for digital transformation in the mortgage sector.
