Smarter AI-powered decisions and automation have become critical elements in loan decisioning today, especially as traditional credit scores wane in importance, says one expert.

Pankaj Jain, president of Scienaptic—which provides AI-powered credit decision platforms––told CUToday.info that when used correctly, AI-driven auto decisioning tools can not only lift performance immediately by pulling in low-hanging, higher-credit-score candidates, it can attract solid borrowers whose scores are unfairly low.

For example, Jain said initial tests among credit unions of its AI-powered credit decisioning platform show potential for 70%-90% increase in the number of automated decision-making opportunities and 20-40% increases in credit approvals for card and auto portfolios.

He stressed these types of numbers are critical to credit unions as fintechs elbow in on the movement’s lending opportunities, and as the lending market tightens as rates rise.

Jain said the gains CUs can make with the use of AI-driven decisioning tools are typically large as credit unions, even in recent years, have relied too much on manual processes.

‘Not Going to Wait’

“Too many credit unions still have too many manual processes in their loan decisioning,” said Jain. “And it is becoming all too clear that the younger members have no patience. If you have someone’s application sitting in the queue and you’re not getting back to them quickly, that’s a big issue. Millennials, especially, are just not going to wait. They will move on to someone else for the loan.”

But the bigger risk, according to Jain, goes beyond losing the loan.

“They may leave for a loan somewhere else, but that experience they just had with you could lead them to have the same opinion about the credit union overall—that it is slow and not what they need in this age where speed, especially for the younger consumers, is so important,” Jain said. “The credit union could lose that entire relationship.”

Jain asserted that AI-driven auto decisioning can immediately lift loan volume by 30% just by moving away from a lending process that relies too much on manual steps.

“That is just the low-hanging fruit, where the system does not have to dig too deeply to determine a borrower’s true risk,” Jain explained.

‘Diamonds in the Rough’

But AI-driven systems can look at thousands of data points quickly to find those good borrowers who would likely be rejected via a manual process, reminded Jain.

“These are the diamonds in the rough that AI can review in less than a second,” said Jain. “There is no reason to miss these opportunities. And you don’t miss them because AI can look at large amounts of data that humans cannot, and find these diamonds that generally require very little follow-up to make the final decision.”

Jain said what currently often happens with lending decisions based too heavily on credit score is that the traditional score is no longer painting a true picture of the borrower. He insisted to get the clear picture of a borrower means looking at many more data credit points, such as debt-to-income, bill payment, rent payment, utility payments and more.

A Plummeting Score

Jain shared that he is aware of one wealthy borrower with a long history of great credit, but since the individual recently took out a loan for a mortgage on an apartment building to take advantage of low rates, the borrower’s FICO score plummeted by more than 200 points.

“That score now does not paint an accurate picture of this borrower who would unfairly be considered a credit risk today if a person or a credit decisioning system just looked at his score,” said Jain, adding that traditional credit scores change too quickly and can be volatile. “You have to look at so many more data points on the borrower and do that quickly. A lot of lenders now have a 680 FICO cutoff, and that is just not always an accurate picture of the borrower anymore.”

Jain, reiterated that too many credit unions still rely on manual processes that are losing loans and members.

“The younger members expect you to perform like Amazon,” said Jain. “AI-driven loan decisioning is here to stay. We have a lot of credit unions in our pipeline, almost 30 now, from large to small. We just got Bethpage FCU up and running and their initial results are right in line with what we had expected—70%-90% increase in the number of automated decision-making opportunities and double-digit increases in credit approvals for card and auto portfolios.”