BW: The rise in covert scams has sparked a global push for stronger customer protection and fraud prevention.
At the 2024 Global Fraud Summit, nations from the Asia Pacific region (APAC), Europe, and North America had endorsed a commitment to combat scams, compensate victims, and punish offenders.
Regulators in APAC are also enhancing fraud prevention, with Hong Kong’s Police Force launching the Anti-Deception Coordination Centre, and Singapore mandating, as part of the country’s Shared Responsibility Framework, real-time fraud detection for banks.
Globally, as consumers seek seamless, personalized transactions, merchants will need to balance security, compliance, and smooth payment processes. Adopting a tailored risk management solution can help businesses maintain security without compromising conversion rates.
DigiconAsia: How do the prevalent payment fraud types in Singapore, such as phishing and refund fraud, compare to those observed in other regions like APAC, Europe, or North America? What practical cybersecurity tips can individuals and businesses worldwide follow to protect against the types of fraud mentioned, regardless of their location?
BW: Countries with similar levels of digital payments maturity, such as Singapore, Europe, and North America, often see similar fraud trends.
With high levels of 3D-Secure adoption and multi-factor authentication, fraudsters have been quick to adapt their tactics.
For instance, phishing attacks may seem like isolated incidents, but successful phishing attacks have a downstream impact as fraudsters make use of stolen details to conduct other forms of fraud such as card testing and account takeovers. This also means that when consumers from these markets fall prey, they may become subsequent victims of more sophisticated scams that are difficult to detect.
Regardless, our experts recommend that businesses globally to look at adopting a multi-layered digital infrastructure that takes into account prevention, detection, and response. Prevention involves analyzing past transactional data to anticipate fraud; detection has to rely on real-time monitoring to flag suspicious activity: and both processes must be paired with a swift response mechanism to minimize impact and protect customer information.
DigiconAsia: What innovative strategies or technologies are being used to combat payment fraud, that could be applied globally? How can lessons learned from tackling scams and creative fraud inform cybersecurity practices for international audiences in regions with varying levels of awareness?
BW: In our experience, there is no silver bullet in fraud prevention: it is all about striking the right balance between security, conversion, and cost.
Risk management has traditionally relied on rules-based risk systems that are often reactive and labor-intensive in approach. However, as fraud techniques evolve rapidly, businesses will need a more adaptive and proactive strategy to stay ahead.
The key is leveraging technology to mitigate fraud before it happens. AI and ML play a crucial role in this mission: by enabling real-time, multi-layered risk management that evolves with emerging threats.
In the case of card testing fraud, machine learning analytics can identify these attempts in real time, allowing businesses to investigate unusual activities such as the increases in issuer refusal rate, or in the number of payment attempts with a $0 amount.
At Adyen, we have seen that leveraging AI and ML in the fight against fraud can reduce manual risk rules* by 86% on average, minimize false positives, and help prevent chargebacks before they occur.
That is why we use risk-based intelligence to differentiate good and bad shoppers, using insights from trillions of dollars’ worth of global payments data on a single platform. This will ensure that genuine customers can check out seamlessly, while fraud detection models are updated continuously to improve security without added friction in transactions.
DigiconAsia thanks Ben Wong for sharing his professional insights with readers.
“which are predefined, human-crafted conditions or criteria set within a fraud prevention system to identify and flag potentially fraudulent transactions. Manual risk rules require human intervention to create, update, and maintain, making them static and often reactive. As fraud tactics evolve rapidly, these rules can become labor-intensive to manage and less effective, prone to missing sophisticated scams or generating false positives that frustrate legitimate customers”
