Skip to Main Content Skip to Footer Content
Close Search

Game-Changing Fraud Detection for South African Financial Institutions

Higher Ed Institutions Can Put Money Back

With fraud on the rise throughout the world, fraud detection must get faster, stronger and above all, smarter. This challenge is compounded by the fact that digital payment volumes, channels and types are all growing. And while machine learning has proven to be an effective ally in combating fraud, today’s financial institutions must go a step further to protect themselves and their customers.

Enter ACI’s new network intelligence approach. ACI fraud expert Cleber Martins and I were recently joined by Raven Naidoo, senior manager of fraud prevention for Nedbank, and Susan Potgieter, head of strategic services for SABRIC, to discuss how a network intelligence approach can be a game-changer in the fight against fraud for financial institutions across South Africa

Fraud challenges in South Africa

Any discussion of today’s fraud environment generally begins with the impact of COVID-19. As noted by SABRIC’s Potgieter during our webinar, fraud levels within South Africa country initially dropped due to the strict lockdown. People weren’t shopping or even using ATMs. But as the world started to re-open, fraudsters returned to take advantage of the rise in eCommerce purchases. According to Nedbank’s Naidoo, card-not-present (CNP) remains the country’s largest contributor to fraud, with the majority being cross-border fraud.

Naidoo also expanded on the rise of phishing scams throughout the country. He mentioned that as day-to-day banking becomes increasingly digital, scammers are able to take advantage through fraud such as authorized push payment (APP) scams. The difficulty here is not just in detection but recovery. Funds sent through this method can be nearly impossible to recover.

How can banks fight back? It starts with behavioral data.

“The crime trends are now in a space where abuse and manipulation of the customer are probably what the perpetrator is aiming to achieve,” said Potgieter. “It’s time for us to bring behavioral signs to the table. It’s not just about using reactive information.”

Defining network intelligence

As discussed in a previous post, machine learning is good, but it isn’t perfect. Machine learning is only as good as the data it receives, making it vulnerable to the siloed consortium structures that dominate today’s payments ecosystem.

This has created a need for a greater intelligence community that can share information in real time. The community is comprised of a larger network of financial institutions (FIs), each producing and sharing more signals than ever before. These signals create a much larger and more detailed picture of the fraud landscape, helping to identify trends and threats faster and more completely.

Ultimately, a transaction will pass through several entities, each revealing more signals to help form the DNA of risk on the transaction. This is in addition to the data elements that FIs are accustomed to receiving and can be consumed immediately by any member on the network. This approach transforms the way signals are made available and their quality, helping FIs make better and more informed decisions.

As Martins explained, “as a financial institution about to send money out, but without the insights from network intelligence, you don’t know anything about the target or where the money is going. But with the network intelligence, we’re really empowering financial institutions to share the reputation of the accounts in real time. So, before the money actually moves, you’re able to get the insight to your machine learning to understand how reliable that account is.”

What the world can learn from South Africa

For Potgieter, the keys to fighting fraud revolve around the core of network intelligence – technology and collaboration. She recommends that banks fully embrace technology and share best practices and information with each other. This three-pronged approach has served the South African market very well.

A key takeaway here is that, in Potgieter’s opinion, fighting crime should not be treated as a competitive differentiator. Banks that seek to win in terms of fraud prevention, ironically, end up making the entire sector more vulnerable. South African financial institutions have been very proactive in working as a collective, including various players such as mobile companies and law enforcement.

Products have also been created to monitor day-to-day activities, leading to timely information. This is one of the ultimate benefits of network intelligence: timeliness. Banks and FIs in South Africa can react quickly to threats, serving to both nullify them in the short term while forcing fraudsters to constantly reinvent their tactics.

Naidoo was quick to point out that shared information allows banks to quickly update their rules strategies and machine learning modules — all in real time. He ends with a short but profound insight: sharing is a requirement nowadays.

 

For more insights from this expert panel, listen to the full “Game-Changing Fraud Detection for South Africa” on-demand webinar.