While instant digital payments translate to better convenience for users, it also means less time for financial institutions to investigate and respond to fraudulent activities. It is inadequate for banks to rely solely on a rules-based detection, and AI and machine learning may offer some reprieve from risks associated to instant payments.

Companies need to swiftly discover more efficient methods for securely enabling remote verification of individuals – in particular identity verification and authentication, two of the most risky activities prone to digital fraud, and which pose significant challenges for businesses navigating digital transformation.

Ian Holmes, Global Lead for Enterprise Fraud Solutions at SAS also explained about two important components – identity and payments – which are vulnerable to being abused by persons to commit crime or fraud.

With finance more digitized than ever, the importance of financial fraud detection and prevention cannot be overstated, but it should not come at the expense of customer experience.

New payment methods offer benefit and immense convenience, like lower cost of financial transactions and wider accessibility/availability for under-banked or unbanked segments, but they also come with vulnerabilities that can be exploited.

Ian added, “This is why the growing magnitude of financial fraud in Malaysia has necessitated governmental intervention such as the doubling of budget allocations to the National Scam Response Centre (NSRC) and the creation of the National Fraud Portal (NFP), a specialized website by the Bank Negara Malaysia (BNM) to gather information about bank accounts and details used by fraudsters.

“Aside from maintaining consumer and business confidence in new payment technologies and ensuring that the economy and society as a whole continue to reap their benefits, fraud detection and prevention in financial services is a must to protect individuals’ finances and businesses’ viability.”

Instant payments and why fraud is happening at these levels now

Ian thinks there are such high levels of fraud due to “…our acceptance of divulging data and inadequate security counterweights.”

The explosion of data in volume and sources have had fraudsters come up with more creative ways to harvest them from customer databases.

“This accessibility to more personally identifiable information has enabled fraudsters to conduct more elaborate and damaging scams, as they can leverage the wealth of information readily available to them, Ian said.

“They use this compromised information to target customers through phishing and smishing and that leads to romance scams, call centre scams and investment scams.  Then, focused attacks through social engineering will increasingly convince the customer to further compromise themselves and to send money directly.”

Widespread use of interconnected devices through the Internet, provide cybercriminals with a plethora of entry points into organizations. Widespread use of generative AI also makes it increasingly challenging to discern between genuine and malicious content like emails and phone calls.

Multi-prong approach to combat fraud

The best strategy for regulators involves adopting a multi-faceted “ecosystem” approach with ‘participation’ from stakeholders as diverse as telcos, traditional banks, newer fintech companies, and of course, consumers themselves. Collaborating and relevant information sharing provides comprehensive and invaluable visibility of upcoming threats. It surfaces emerging patterns of financial fraud that can be identified more quickly than before.

Financial institutions can now run thousands of iterations using different and more refined algorithms, leading to improved fraud detection while enhancing the overall digital customer experience.

Consumer education is the first line of defense. Comprehensive awareness campaigns about common online fraud tactics, protective measures, and reporting channels can empower consumers to detect and report fraudulent activities promptly.

By also maintaining a dynamic approach, regulators can continuously refine their strategies and ensure that they are equipped to tackle new and evolving forms of online scams.

With the number of face-to-face transactions reducing, effective fraud detection in payments requires immediate transaction data to be supplemented by additional contextual data such as information about the device being used, its historical usage, biometric identification data, and the user’s patterns of activity of the user.

Ian said, “By analyzing these multiple inputs, advanced data analytics solutions can establish a baseline of what is considered normal activity and promptly identify and investigate any suspicious activities.”

With AI techniques, large volumes of structured and unstructured data can be analyzed to identify patterns and anomalies indicative of fraudulent activities, and then also complemented with machine learning models that continuously learn from historical data and update their fraud detection algorithms in real-time.  Financial institutions can now run thousands of iterations using different and more refined algorithms, leading to improved fraud detection while enhancing the overall digital customer experience.