A smaller share of mortgage lenders in 2023 say they have deployed artificial intelligence (AI) and machine learning (ML) within their business operations compared to five years ago, according to a recent report from Fannie Mae.
The findings come from a survey of 242 senior executives representing 219 lending institutions nationwide. The survey, which specifically inquires about business practices tied to AI and ML within the mortgage industry, is a new iteration of one that Fannie Mae’s Economic and Strategic Research Group performed in 2018.
Familiarity with AI and ML has stayed consistent over the past five years, with almost two-thirds of mortgage executives saying they are “very familiar” or “somewhat familiar” with the technology. Sixty-three percent of survey respondents indicated some level of familiarity in 2018, with that share growing by two percentage points this year.
But only 7% of this year’s survey respondents said that their company has deployed AI solutions into its current mortgage process. Interestingly, that’s down substantially from a 14% share in 2018. But 22% of respondents said that their companies have begun to deploy AI and ML solutions on a limited or trial basis, up 9 percentage points in five years. Taken together, 30% reported having deployed AI technology on at least a limited basis in 2023, compared to 27% in 2018.
Forty-one percent said their companies were “investigating” AI and ML solutions but have yet to utilize them. That’s up from 36% in 2018.
Notably, the focus of AI adoption among mortgage companies appears to have solidified toward the improvement of operational efficiency. Some 42% of respondents in 2018 said that improving operational efficiency was their firm’s primary objective in pursuing AI adoption, the largest share among all responses. But in 2023, this share skyrocketed to 73%. Meanwhile, enhancing the borrower experience, which garnered 41% of the share in 2018, plummeted to only 7% this year.
When it comes to impediments to AI adoption, most survey participants cited the difficulty of assimilating the nascent tech with their current systems, mirroring results from the 2018 survey. In 2018, 50% said that the complexity of integrating AI/ML apps within their existing infrastructure was either the largest or second-largest challenge involved in implementation. That share decreased by 2 percentage points this year, although it still received the most responses among a list of barriers to AI adoption.
The lack of a proven track record of success garnered the second-most responses in 2023, with 35% citing it as the largest or second-largest challenge, followed by high costs at 24%. Concerns with data security and privacy jumped from 10% of responses in 2018 to 22% this year as lenders and other financial institutions continue to grapple with the challenge of keeping sensitive consumer information secure from increasingly sophisticated digital attacks.
