As peer-to-peer (P2P) payments become increasingly popular, financial institutions face growing challenges in identifying fraudulent transactions that appear to be authorized by legitimate account holders. Unlike traditional payment fraud, authorized push payment (APP) scams rely on social engineering to convince consumers to willingly transfer money to criminals, making these transactions particularly difficult to detect using conventional fraud prevention methods.
FICO’s introduction of behavioral analysis capabilities within its Falcon Fraud Manager reflects the industry’s growing emphasis on machine learning, behavioral biometrics, and advanced fraud detection. However, fintech companies such as BioCatch, NuData, and ThreatMetrix have already spent years developing similar technologies. The question for financial institutions is whether FICO’s solution can deliver comparable detection accuracy, operational efficiency, and value.
FICO Expands Fraud Detection With Behavioral Analytics
FICO announced a new solution in FICO Fraud Manager that utilizes behavioral analysis and other signals to prevent P2P fraud. Fintechs have been doing this for years.
Fintech Companies Bring Established Fraud Prevention Capabilities
Mastercard acquired NuData, LexiNexis acquired Threatmetrix, while BioCatch and others continue to go it alone. The Fintechs have been in market for several years and have used that time to hone their machine learning algorithms.
For example, BioCatch claims to detect 5 different unique forms of criminal mule activity that impact accounts. It will be interesting to see if the FICO solution is equal to the Fintech solutions from the perspective of price/performance:
Detecting Authorized Push Payment Scams
“Alternatively, should a consumer use her bank’s mobile app on her own phone, but sends funds to a new account, the likelihood is 10 times greater that she is falling victim to an APP scam, Zoldi adds. When it comes to consumer’s favorite devices, the Scam Detection Score identifies 24 times more scams than the standard fraud score, FICO says.
What makes APP scams more difficult to detect is that they use social-engineering techniques to trick consumers into sending money from a personal account to an account controlled by the criminal for what consumers believe is a legitimate reason. “This means that the model must look for subtle patterns that point to … what legitimate customers do when being misled by criminals,” Zoldi says. “The typical hallmarks of third-party fraud that look out-of-pattern don’t necessarily exist for APP scams.”
Social Engineering Makes Payment Fraud Harder to Identify
Criminals enacting a push-payment scam may reach out to victims through mobile games, online shopping sites, and social media. Online gaming users, for example, may believe they are paying for a rare item. Or online shoppers may believe they are buying a legitimate product. With social-media scams, criminals have been known to spend months grooming victims through online conversations, developing a relationship with the target before asking for money to deal with a fictional emergency.
“Whatever the platform, victims believe they are receiving a legitimate service, product, or benefit,” Zoldi says.”
The introduction of behavioral analytics into FICO’s fraud management capabilities highlights the growing importance of detecting suspicious activity beyond traditional transaction patterns. As criminals increasingly manipulate legitimate account holders into authorizing payments, financial institutions need fraud prevention technologies capable of recognizing subtle behavioral indicators that conventional fraud scoring systems may overlook.
Behavioral biometrics and machine learning offer potential advantages by examining how consumers interact with their devices, navigate applications, and initiate transactions. These signals can help identify unusual activity even when a payment originates from a recognized device or authenticated account. For authorized push payment scams, this additional context is particularly valuable because the transaction itself may appear entirely legitimate.
However, the competitive landscape presents challenges for established fraud prevention providers. Fintech companies specializing in behavioral analytics have spent years collecting data, refining detection models, and identifying emerging fraud patterns. Their experience may provide advantages in recognizing sophisticated scams, including money mule activity and social engineering attacks. FICO’s ability to compete will depend not only on detection performance but also on implementation requirements, pricing, and the effectiveness of its technology in reducing false positives.
For banks and payment providers, evaluating these solutions requires balancing fraud prevention with customer experience. Excessive transaction declines or unnecessary authentication challenges can frustrate legitimate customers, while insufficient controls expose institutions and consumers to financial losses. As authorized push payment fraud continues to evolve, financial institutions will increasingly need layered detection strategies that combine behavioral intelligence, transaction monitoring, and machine learning to identify suspicious payments before funds leave customer accounts.
Overview by Tim Sloane, VP, Payments Innovation at Mercator Advisory Group
