The payments ecosystem is evolving at a remarkable pace. Unfortunately, so is fraud. The threats are pervasive, intelligent and responsive, reacting to digital innovations as quickly as they appear. Financial institutions are under pressure to respond rapidly and with solutions that address the multi-layered approaches taken by criminals. As Damon Madden, regional solution consultant and fraud expert at ACI Worldwide, points out, the threats use personal and psychological levers alongside advanced technology to manipulate customers effectively.
“You’re not just looking at compromised credentials anymore,” he says. “What we’re seeing now are highly sophisticated scams built on social engineering where customers are manipulated into making the payments themselves.”
These scams, referred to as authorised push payment fraud, are among the hardest to detect. Whether it’s a romance scam stretched over several months or an urgent call from someone impersonating SARS or law enforcement, the tactics are targeted and emotional. Unfortunately, they’re also very effective. “Criminals are deploying AI and machine learning just as financial institutions are,” says Madden. “As the banks get smarter and faster, so do the criminals.”
SA is particularly vulnerable. With high levels of mobile penetration and varied levels of financial literacy, the market is an attractive target for well-organised fraud networks. USSD payments, for example, are still widely used in the country but they are susceptible to fraud as they have weak identification mechanisms. The criminals are using tools and platforms like Telegram and WhatsApp to perpetrate effective fraud campaigns that are built on the illusion of trust and belonging.
Fraud prevention tools designed to combat these increasingly pervasive and intelligent threats are essential. Financial institutions also have to stop thinking of fraud defence as a back-office function.
“Customers are choosing their financial providers based on how well their finances are protected,” says Yvonne Elford, sales executive at ACI Worldwide. “That’s a competitive differentiator. Customers want to trust in the service and will choose providers that prioritise that trust. We’re working closely with banks to help them protect transactions and relationships with prevention that builds confidence and loyalty.”
That said, legacy systems and siloed data remain barriers for many companies. “Fraud tools are operated in isolation, and this makes it harder for companies to build a full view of the customer journey. You need an enterprise-level perspective designed to detect what’s legitimate and what’s not, without creating friction for genuine users.”
This is where behavioural biometrics, device intelligence and real-time profiling come in. Madden describes how adaptive authentication and scoring systems, powered by machine learning, can analyse essential metrics like hesitation, changes in device usage or unexpected beneficiary patterns, all without disrupting the experience.
“It’s about protecting the user without obstructing their experience,” he says. “The goal is to reduce false positives while keeping the fraudsters out and allowing your customers to transact smoothly. However, AI won’t solve everything. You need real experience in payments, a deep understanding of the data and scalable models capable of handling thousands of transactions per second.”
What works, he says, is targeted machine learning designed to reduce alert fatigue, streamline compliance and help analysts focus on what matters. “We’ve seen up to a 60% reduction in rule volume by shifting to a scoring-based approach. But, more importantly, we’re giving fraud managers the ability to adapt. To write their own rules, react to changing trends and integrate with other intelligence providers in real-time.”
* ITWeb TV Biz episodes are sponsored.
Share