Ken Jochims Sep 9, 2020 5 min read

Top considerations for behavioral biometrics solutions

Revelock recently took part in a webinar with our partner, NICE Actimize, to discuss how they were incorporating our behavioral biometric anti-fraud technology into their online security systems to help prevent online banking fraud.

As discussed in the webinar, although transaction amounts among legitimate users have remained relatively stable in recent years, the amounts lost as a result of mobile and online banking fraud have steadily increased. It is clear that security around online banking transactions needs to be increased to counteract this trend. This is where Revelock

behavioral biometrics steps in to help.

During the webinar, listeners voted in a live poll on their top criteria for a behavioral biometrics-based anti-fraud solution. The results of this can be seen in the image below:

behavioral-biometric-accuracy-consistencyAs we can see, the results point overwhelmingly towards the accuracy of such a solution, and the consistency that can be achieved across different users.

These are understandable concerns to have. If the solution in question cannot identify users to the highest degree of accuracy and consistency, the probability of false positives and false negatives occurring is very high.

This would likely result in legitimate customers being asked to complete unnecessary extra checks, which hinders the frictionless experience customers have come to expect from online banking and online payments.

And all the while actual fraudsters are slipping through the net.

By analyzing behavioral biometrics alongside other contextual information about the user and their transactions, Revlock’s solution delivers better detection and improved alert quality, which only challenges a user if all this information combined generates a level of risk high enough.

Additionally, these false positives and negatives are often generated because users are compared with groups, or ‘clusters’, of users. The more sophisticated fraudsters who are familiar with application layouts or may have even studied user behavior through online monitoring tools have a higher chance of successfully perpetrating fraud and stealing funds.

Plus, if an individual user is not behaving in a way similar to the ‘community’, for example, if they travel often for work and so regularly log on from different IP addresses and geolocations, there is a high risk of a false positive being generated each time they log in.

Revlock’s solution is extremely specific, profiling each user individually and employing AI to compare a user’s online interactions against their entire personal history. Alongside this, the solution’s deep learning technology means the solution becomes increasingly fine-tuned, edging closer to 100 percent accuracy every time a user logs in.


Watch full webinar


You might be interested in our post Physical Biometrics vs Behavioral Biometrics


Ken Jochims

Ken has over 25 years of enterprise software product marketing experience delivering fraud prevention, customer support, identity and access management and IT infrastructure solutions to financial institutions and fortune 1000 companies. Prior to Arxan Technology Ken worked for Neustar, ThreatMetrix, Guardian Analytics, Genesys, CA Technologies, NeXT Computer and Apple. Ken received a BS in Engineering Technology from California State University, Long Beach, and outside of work Ken can be found hiking, mountain biking and working on cars.