CO-OP, PSCU Advance Fraud-Detection Systems

Two industry heavyweights are moving forward with new fraud-detection systems designed to root out suspicious activity by capitalizing on the growing prowess of artificial intelligence, machine learning and data analytics.

CO-OP Financial Services announced last week that it began a pilot test of its COOPER Fraud Analyzer system, and PSCU, which is a payments CUSO headquartered in St. Petersburg, Fla., announced an “exponential month-to-month increase” in fraud blocked by its new Linked Analysis system.

CO-OP, based in Rancho Cucamonga, Calif., first unveiled its COOPER Fraud System in May. As part of its new pilot test, the system funnels shared branch account-based transactions from four credit unions through a system that scans for and flags potentially fraudulent activity.

CO-OP said it plans to make the system available to more than 1,800 credit unions in the shared-branch network after the pilot testing is complete. In 2019, it plans to launch a fraud-scoring model that uses machine learning to rate the level of suspicion for transactions.

The CO-OP network includes more than 5,700 locations across the country, making it the second largest system of financial services branches in the United States, CO-OP Financial Services said.

“It is CO-OP’s job to make sure that our clients have access to the best technology to combat fraud and protect their members,” CO-OP President and CEO Todd Clark said. “We believe the most effective answer lies in developing a machine learning and artificial intelligence solution that we can bring to our credit unions at scale. The launch of COOPER Fraud Analyzer into client testing marks a major milestone in the evolution of CO-OP as a financial technology provider, and it is particularly significant in that we are first providing this tool to one of the industry’s most valuable assets, the CO-OP Shared Branch network.”

PSCU’s new Linked Analysis system, launched in April, looks at members’ activity across different platforms, different institutions and merchants across any card. Data scientists and machine learning use the data to identify linkages, compromises, breaches and common points of purchases.

The system detects activity that’s inconsistent with members’ typical patterns and then blocks cards before fraud attempts occur; it also monitors for suspicious circumstances such as multiple failed authentication attempts on different platforms. The idea is to predict and mitigate fraud, the CUSO said.

Over the last year, PSCU was able to stop more than $200 million in fraudulent transactions, it said.

“By using shared intelligence to predict and defend against fraud, we are enabling a greater member experience for our owners,” PSCU Chief Risk Officer Jack Lynch said.

“Fraud has changed dramatically over the past couple of years…We are now looking at all of the data available to us, identifying where there might be potential trends in the future, and using that information to prevent and detect fraud sooner,” he added.