Digital transformation (DX) has been changing the face of financial services across the globe by enabling higher agility, customer satisfaction, and business growth. Digital payments are witnessing a massive increase, with an estimation of double-digit growth by 2022, as per the World Payments Report 2019.  However, while going digital has made it easier for both financial institutions (FIs) as well as customers to carry out financial transactions, it has also opened doors to an increased risk of frauds. The rapid expansion of digital channels, exploding growth in number and types of devices and reduced customer ‘face time’ are making FIs more vulnerable to online fraudulent activities. From 2015 to 2018, there has been a 575% increase in online synthetic identity fraud of which, 49% of all risky transactions stem from mobile devices, as mobile surpasses desktop in popularity.

‍

iovation Financial Services Fraud and Consumer Trust Report 2019 states that includes findings from 5.9 billion financial services and 1,604 consumer interviews both in the U.S. and the U.K. states that in the first half of 2019, iovation saw 50% of risky transactions originating from mobile devices, up from 41% in 2018, and 21% in 2017. 39% of consumers surveyed have closed a financial services account due to security and fraud concerns. Meanwhile, 76% report the most important factor when choosing a financial services account is security and privacy.

‍

Another study by KPMG - Global Banking Fraud Survey 2019 - revealed that over 50% of respondents globally experienced an increase in fraud value, while 60% of respondents witnessed an increased in fraud volume. Further, the survey highlights “evolving digital channels and faster payment processing” as one of the major challenges of today’s banking system.

‍

AI and ML-based solutions: Preventing fraud-risk in the digital era

While fraud prevention is undoubtedly the need of the hour, an overly aggressive fraud detection mechanism may harm the customer experience by slowing transaction speed, requiring customers to perform too many steps before checkout, or worse flagging good consumers as fraudsters or suspicious. Banks and FIs need smarter fraud prevention that is real time, accurate, and quick. They are therefore increasingly turning towards Artificial Intelligence (AI) and Machine Learning (ML) based solutions that make it possible to detect frauds and suspicious activities in real-time by scanning through humongous volumes of external and internal data. As consumers increasingly leave their digital footprint in the mobile world, smartphone metadata has emerged as a powerful tool in alternative credit scoring mechanisms and fraud detection.

‍

‍MasterCard was one of the first-movers to deploy innovative tech solutions for fraud management and was therefore able to bring down the false declines to its customers by as much as 80%. Leading AI-based solution provider CredoLab that leverages smartphone metadata to build digital credit scorecards has therefore joined hands with iovation, a TransUnion company, specializing in sharpening device intelligence. iovation’s algorithm encompasses 45 application fraud checks, home and work address verification, employment verification, KYC documents collection and alternative credit scoring.

‍

Together, CredoLab and iovation help financial firms clearly distinguish fraudsters from good consumers based on device recognition, context and reputation. This partnership drives further CredoLab’s mission to support better credit decisions by enabling device-based fraud detection and authentication solutions to help banks and lenders to become secure against financial frauds such as identity theft/new account fraud, synthetic identity fraud, or account takeover fraud.

‍

• Quick and accurate customer verification: Banks and FIs need a one-stop-shop solution to evaluate potential customers in a secure, transparent and cost-effective manner.
• Credit access to unbanked, new-to-credit or any other applicant with a "thin file": Advanced data analytics and AI-embedded credit scoring systems help banks and FIs not only in seamless customer on-boarding but also in verifying various customer information attributes to generate and maintain credit scores of individuals outside the traditional credit ecosystem. This helps lenders identify potential customers and their creditworthiness more credibly by converting their ‘thin’ credit history files into ‘thick’ ones, thereby improving overall financial inclusion.
• Tracking of multiple types of frauds: Device recognition and intelligence technology is increasingly being adopted by FIs to detect different types of frauds using attributes such as geolocation, anomalies, device-usage patterns and other indicators to analyze customer profiles, behavior and intent. These indicators are studied to recognize suspicious behavior to prevent and fraudulent occurrences.

‍

Combine supervised and unsupervised Machine Learning to stay a step ahead

While adopting supervised ML first is easier for many organizations (most have resources well versed in foundational AI and ML concepts are now familiar to most banks and FIs need an integrated solution that combines supervised and unsupervised ML insights to generate a single fraud risk score, given the nature of their business. Leaders in this field have the capability to generate a twice as accurate score in a matter of milliseconds as compared to traditional AI approaches that rely on rules and predictive models. Clearly, it’s time to innovate to bear fraudsters at their own game – are you prepared?

‍

Read more about CredoLab and iovation's joint solution here.

‍