The retail banking industry is facing an epidemic of fraud, from lower-level account compromise fraud to sophisticated Authorized Push Payment (APP) schemes. The magnitude of this problem is widely recognized in the industry, and many banks and payment processors are trying to combat it with new fraud-prevention tools and processes.
Unfortunately, fraud is constantly evolving, and most tools can’t keep up with detecting emerging threats. Even when banks detect APP fraud activity, it's often too late, or they lack the data and visibility to alert customers quickly enough—if at all. The result is honest customers losing millions of dollars every year.
However, a new generation of data analysis tools and real-time event streams are changing the dynamics of fraud prevention. These tools build on innovative behavioral and technological insights, machine learning, artificial intelligence, and big data to crunch vast amounts of data in milliseconds and provide fraud prevention teams with robust, actionable, real-time data.
The basic premise of APP fraud is that fraudsters use a variety of manipulative means to con people into sending them money via online banking—specifically, direct payments. Unfortunately, fraudsters have found that preying upon and manipulating real customers is one of the most effective methods to make money. And because APP fraud is one of the most challenging types of fraud to detect, it's rapidly growing in popularity, variety, and complexity. Zelle is a perfect example of APP and is exploited by many fraudsters due to its ease of use and low-security measures.
APP fraud is an umbrella term, with a variety of fraud sub-types incorporated into it. Here’s a rundown on three of the most common (and growing) types of APP fraud:
Perhaps one of the most deceitful forms of APP fraud is romance fraud. This fraud manipulates the human desire for love, friendship, and relationship. In romance fraud, fraudsters identify a potential target somewhere on the web (for example, a dating website) and create fake accounts to try to start a fake relationship with the victim.
If they successfully make contact, they'll slowly cultivate a relationship until the victim trusts the fraudster. The fraudsters eventually make a small request for money. If this works, the fraudsters will make more significant moves to con the victim out of larger and larger sums of money.
The sad part is when the victim finally realizes that not only is the relationship a complete sham, they've also lost a significant amount of money, perhaps even their life savings. Romance fraud increased 80% from 2020 to 2021 alone, according to the FTC.
Another common type of APP fraud is investment fraud, which preys on the human inclination to look for a good deal. This type of fraud often involves a fraudster presenting an investment opportunity that's too good to be true. Cryptocurrency scams top the list of rising investment fraud.
Sometimes fraudsters will even impersonate a trusted friend or contact to add an increased sense of obligation to spend or invest. When the victim finally acts and "invests," they don't even realize they're sending money directly to the fraudsters. In the UK alone, the total amount lost to investment fraud rose 49.58% in the 2021/2022 fiscal year.
By far, the most common APP fraud is impersonation fraud. Though similar to the above types of fraud, impersonation fraud generally refers to various fraudulent activities involving a fraudster claiming to be someone they're not in an attempt to con people out of money.
Suppose you receive an unexpected robocall claiming you owe money to the IRS. Most likely, you're the target of an impersonation fraud attack. COVID-19 also prompted an uptick in impersonation fraud as criminals try to steal employment benefits and funds from relief programs.
You probably won’t fall for the most obvious attempts at impersonation fraud, but fraudsters are becoming increasingly clever at finding new ways to trick even the savviest people.
The simple question is, how do you prevent this type of fraud when the person transferring the money is a real, authenticated customer unwittingly stealing from him or herself? The answer is comprehensive data capture. Leveraging real-time event streams and data analytics rise to meet the challenge of preventing APP fraud by:
Good data is foundational to combating any type of fraud, and the same holds true with APP fraud. You should be collecting and analyzing as much data as you can about every customer interaction on your website or app, including technical data, interaction data, and behavioral biometrics. A modern data capture solution collects every insight available, such as mouse movements, payment requests, session details, IP addresses, device information, time spent per page, etc. to create a comprehensive identity graph for every individual.
One of the key insights powering modern fraud prevention is defining what "normal" is for any given customer. For example, if a user who never transfers money electronically suddenly initiates a large transfer, you may have fraud in progress. The only way you know this behavior is abnormal is by collecting and analyzing behavioral data to paint a picture of regular activity for a particular user. From this starting point, deviations from the norm are easier to detect.
The identity graph becomes the benchmark and every interaction is then compared against it to detect anomalies. Leveraging first-party data, AI, and machine learning means the data set is continually growing over time to evolve the identity graph.
You can then gather actionable insights from your data. For example, did you know it’s possible to detect emotional duress through erratic mouse movements? When a person feels scared, uncertain, or threatened, they unconsciously show signs of duress which can be detected in their behavior. Fraudsters are notorious for using a sense of fear or urgency to manipulate victims into making decisions quickly before they have a chance to think. These subtle cues are often only detectable by algorithms and machine learning. But they're detectable nonetheless!
Capturing and analyzing all this data isn’t enough. Most fraud solutions can’t put any of the triggers or actions in place in real time. At best, they can provide a basic score with no context. If you’re only able to act on the data after the fact, it’s too late. The true magic in preventing APP fraud is to feed all of this technical and behavioral data into your fraud prevention tools in real time.
With a modern data capture solution, real-time event streams funnel the various technical and behavioral events to the right place, analyzing and crunching them as they occur. When a suspicious event (or events) trigger, your systems and teams are immediately notified and provided the information they need. Your fraud prevention team or system can then take action to intervene, either blocking the transaction or contacting the customer to review the situation
The difference is in the ability to take action in real time. Having a score, or even the data, in real time is irrelevant if you can’t act on it instantly. A comprehensive data capture solution that employs real-time event streams increases your ability to intercept fraud at the point of execution, rather than simply cleaning up the mess after the fact.
This early intervention inevitably reduces fraud losses for the bank and their customers, builds customer loyalty, and protects them from getting conned. And as a bonus, stopping fraud before it happens also saves your teams countless hours on the phone helping duped customers clean up the mess.
Simply put, it's a win-win-win.