The latest version of the RiskOps Platform features scam-specific machine learning (ML) models together with other capabilities that look to deliver detection, classification, and interception of an extensive array of payment scams and financial fraud typologies.
The announcement details that financial scams are an epidemic growing in both dollar volume and human impact, and the FTC has reported that between 2021-2023, losses due to scams increased by 30% in the US. These losses are only at the surface level, as globally, 7% of scam losses are reported, as per Global Anti-Scam Alliance (GASA) data. Additionally, as scammers can access generative AI using large language models (LLMs) to create increasingly sophisticated and realistic scams, the losses are expected to further increase.
Furthermore, most FIs employ multiple legacy point solutions that leverage rules-based approaches for the identification of fraud types but fail to target scams specifically, as these approaches have three critical limitations. Per the press release, they are siloed by channel, making them susceptible to fraud typologies operating across a multitude of banking products or payment platforms. They identify fraud by focusing on either behavioural activity or financial activity but do not visualise them both together, which limits the ability to diagnose a scam in progress rapidly. Additionally, such approaches fail to evolve quickly to protect against new fraud types as scammers constantly change tactics.
Improved AI-based detection, with the improvements to Feedzai’s AI engine, based on learnings from multiple geographies and FIs, set to enable better scam detection with decreased false positives.
Custom scam rules, capability for adding bank-specific scam detection rules and thresholds in addition to out-of-the-box AI detection to deliver a quick and customised response to new attacks.
Early intervention. Digital signals, including behavioural biometric and non-transactional patterns, could suggest a customer might fall victim to scams ahead of them making the payment, and Feedzai’s features enable banks to identify such signals and create both intervention and education opportunities earlier in the risk lifecycle.
Inbound payments. For a scam to benefit them, scammers need to send money to accounts they can control. The company’s features help monitor inbound payments together with outbound ones, providing banks with multiple opportunities for stopping scams.
Extended scam reporting, which includes customisable scam classification, performance metrics, and reporting.
Commenting on the launch, Pedro Barata, Chief Product Officer of Feedzai said that consumer losses due to scams are expected to increase as countries launch or expand instant payment channels, exemplifying the US which is to introduce FedNow. Scams’ impact on humans is especially high as bad actors predominantly target vulnerable groups like the elderly, those at an economic disadvantage, immigrants, and youth. Per their statement, as it becomes increasingly fast and easy to make payments, new solutions that enable scam detection and interception ahead of money moving are required, adding that the ScamProtect features look to provide a ‘comprehensive approach’ to scam prevention.
Every day we send out a free e-mail with the most important headlines of the last 24 hours.
Subscribe now