Article

Buttoning Down Compliance Using Automation

blue button on a blue shirt
Contributing Writer
member of Bellco Credit Union

3 minutes

The power and limits of using technology to meet regulatory requirements

Regulatory compliance monitoring can be automated reliably up to a point. Technology has made it far easier for most credit unions to button down the compliance required around onboarding members and initiating loans, notes Marcus King, VP/audit and compliance for Credit Union Audit and Compliance Group, Birmingham, Alabama.

“The software for those activities will alert a staff person or an online program when a compliance fault occurs—and even prevent the process from continuing until it is fixed,” he explains. Most CUs have this technology, he adds.

The rewards of automation can be huge, especially around high-volume transactions. Regulations require CUs to have strong fraud detection and prevention systems. Automation tools can study patterns in members’ payments and create templates that detect “suspicious” (out-of-the-ordinary) payments and trigger alerts for people to process.  

The problem, according to Amber Sutherland, SVP/sales at Silent Eight, a regulatory technology company with locations around the world, is that you get a lot of false positives that still require too much staff time to investigate. But artificial intelligence, she claims, can automatically apply more data and more sophisticated analysis to shrink dramatically the number of alerts that are really suspicious.
               
The same is true for compliance with “know-your-customer” regulations that require CUs to onboard only people who can be verified as legitimate applicants. Onboarding volume is up since COVID-29, Sutherland reports. Tapping multiple databases in seconds, AI can accurately separate real people from fraudsters in many cases, she explains.

For something as basic as automated onboarding or loan originations, software can make sure all required fields are populated in a compliant manner, notes Bryan Boynar, financial services solution marketing manager for CUESolutions provider Hyland Software, Westlake, Ohio. And it can confirm that the information squares with other information about that member in available databases. That addresses preventable problems, he says, and does so before the onboarding or lending process can proceed—without creating flags or alerts that humans would have to investigate.

Automation plays a big role in compliance and has for years, Gaye DeCesare points out. She is VP/compliance for Aux, a compliance-focused CUSO based in Denver. You can set strategic parameters in computer programs and let the macros lead staff or members through application processes, even halt their progress until they get it right, she explains. Then the computer can screen against the parameters and spit out exceptions for staff to address. Automation can prompt staff for what they should do and what needs to be reported.

“It’s basically necessary to let the computer screen for suspicious transactions,” she says. “It’s too big a job to do manually.”
               
But then human eyes must take over.

“You can’t set it and forget it,” DeCesare insists. “Unusual activity is not always illegal and illegal activity doesn’t always look unusual. You have to do periodic validations of the automation and updates if you add a service or member group.” And human intelligence has to take over when a member needs to be contacted.

“Automation can be quick, accurate and efficient, but it’s not good at understanding,” she concludes.

Automation can shine, observes CUES member Dale Scott, CCE director, risk services, at the $5 billion ConnectFirst Credit Union, Calgary, Alberta, especially for the black-and-white issues where something is or is not compliant. This can be used in some instances with new online membership and account opening as well as new loan applications. A member using an online channel who hits a compliance snag, he explains, can click on a pop-up box that will explain the issue and what they probably need to do. What can’t be automated effectively are the gray areas in which a member’s character or specific circumstances come into play—and CUs shouldn’t want to automate that activity, he insists.

“That’s where we shine, what makes us special,” he points out.

Richard H. Gamble writes from Grand Junction, Colorado.

CUES Learning Portal