Dyre Straights: How to Best Combat the Latest Generation of Malware

The information security professional’s worst nightmare is the cybercriminal attack imposter armed with a legitimate user’s valid credentials. Unfortunately, the cybercriminals are adept at stealing them. Valid usernames and passwords can be lost in data breaches, keystroke loggers may capture them and ship them off, an end user can be socially engineered to reveal them – the list is long. These examples, however, are only bad dreams compared the latest generation of malware variants called Dyre or Dyreza – a Remote Access Trojan, or RAT.

Computer security researchers have seen a spike in the distribution of Dyre and a significant increase in the sophistication of attacks it can launch. The Dyre banking malware specifically targets user account credentials, and according to Kaspersky Lab, the malware has been stealing millions of dollars from corporate bank accounts by using a mixture of Trojans and social engineering. A particularly onerous variant is nicknamed the Dyre Wolf. This exploit combines remote access, Man-in-the-Browser (MITB) manipulation, social engineering and finally distributed denial of service attacks (DDoS) to distract a user from discovering they’ve been victimized. Dyre Wolf is capable of detecting logins, capturing a user’s credentials, and launching an attack against more than 1,000 global financial services firms.

Cybercriminals armed with valid credentials are difficult to defend against. Inserting additional authentication steps and stronger security at the “front door” will not be able to stop the Dyre attack for one reason: This session looks like it’s a perfectly normal session. Nor does the Trojan respond to any baiting attempt because it doesn’t do anything at this stage. The PC is just remotely controlled. There is no way for the victim or the banking site to know that.

In effect, layers of defense based on geolocation, device signature and proxy detection are all muted by RATs like Dyre or Dyre Wolf, as the attacks are launched remotely from the end user’s own computer.

Biometrics of various types have been suggested as a solution to the RAT problem. Biometrics can ensure the end user is the same end user and stop some attacks, but in the case of Dyre Wolf, the attack is launched from within the user’s authenticated session. So how does one spot the imposter, the wolf in the user’s clothing?

Cognitive or behavioral biometrics solve this problem. This new type of biometrics, also referred to as Next-Generation Biometrics, relate to the way the user is behaving with an application or a site and the habits and preferences he is choosing while online. Furthermore, each user, fraudster or even a malware is profiled in a unique way that is being updated thanks to a machine-learning algorithm.

Different human end users vary in the way they navigate an application. They type differently, wield their mouse differently, and navigate through a website differently. These differences in behavior can be measured, analyzed and modeled as effectively as traditional biometrics. RAT-powered bots behave very uniformly and have their own consistent behavioral fluency - a ‘signature’ if you will -which can be recognized when measuring these types of behaviors. In the case of the Dyre detection, BioCatch can identify the fact that the user is behaving suspiciously by using a remote access tool.

Futuristic? Perhaps, but attacks such as Dyre Wolf are growing on a daily basis. Thus, spotting malware in the act by identifying non-human behavior is today the only way to actually identify a Dyre malware attack.

Early Warning in a strategic alliance with BioCatch has pioneered the use of these types of cognitive measures and behavioral biometrics for successfully protecting user accounts from Dyre and other RATs. When sneakier RATs attack, it’s time to get a better RAT trap.

About the Authors

Oren Kedem is Vice President of Product Management at BioCatch, a Cognitive Biometric™, Authentication and Malware Detection solution for mobile and web applications.

Michael Toth is Vice President of Early Warning’s Product Management Mobile Solutions Team.