RANCHO-CUCAMONGA, Calif.—CO-OP Financial Services will be leveraging artificial intelligence (AI) in the new year to improve performance of the fraud solutions it provides credit unions.
The CUSO sees AI as not only enhancing the ability of fraud tools to detect and prevent threats, but to also reduce the number of “false positives.”
“We will be implementing machine learning tools into our fraud solutions in 2017,” said Amanda Smith, VP of account systems and support. “We will be putting in artificial intelligence platforms to help with better predictability. AI will help us adapt more quickly to changing fraud patterns and allow us to iterate more quickly on models we write.”
Smith’s news comes as part of a series being published by CUToday.info that examines how artificial intelligence, often referred to as “AI,” will manifest itself at CUs.
False Positives
As financial institutions clamp down on fraud, the number of false positives—legitimate card purchases flagged as fraud—has become an increasing problem.
Smith said that while CO-OP’s solutions have always performed well in limiting the number of false positives, employing artificial intelligence should shrink that number even further.
“With AI what we are doing is fine tuning fraud predictability to catch instances that are truly fraud,” said Smith.
Smith said what AI brings to fraud solutions and fraud modeling is the ability to crunch massive amounts of data to come up with better conclusions than humans.
“It’s just a massive amount of data that needs to be churned through to find fraud patterns, recognize changes in behaviors . . . A person, or a team, simply can’t do that as effectively as AI. The more data points you can churn through the more meaningful the results.”
What is also helping pull artificial intelligence into fraud solutions is that the tools for managing large amounts of data have become more affordable, added Smith.
As far as interacting with members, Smith said she thinks AI will first make its appearance in CUs via member service and staff support functions.
“Things like chatbots and ways to assist staff in their call center activity,” Smith said.
Robo-Advisors
Smith also sees AI being used as a “robo-advisor,” a financial advisor that a member could contact via telephone or online channels to address basic banking questions. Some large banks are currently piloting this application for AI.
“I believe this is a strong application for artificial intelligence, as a way to reduce member support costs by pushing some member support functionality to an AI platform,” said Smith, adding that we may see some CUs use AI in this manner in 2017.
Smith predicts that 2017 is still early for AI to make an appearance inside credit union branches in some physical form, such as a robot greeter.
“I see artificial intelligence used more for back-end applications—the things I have mentioned, and maybe to augment risk modeling to benefit lending—than physical robots that look like humans, wandering around in branches—although that might happen someday,” said Smith.
“This is amazing technology for sure,” concluded Smith. “I saw a report other day that showed one of these robots that looked just like a person. But AI is in its early stages, and how it will be used directly with members remains to be seen. I think it will depend a lot on how people react to it. Again, I do see a lot of use cases for AI on the back end with credit unions. And putting robots in the lobby, for example, may be trying to solve a problem that does not exist.”