Protecting your organization against fraud is a continuous game of cat and mouse. It seems like as soon as you implement a detection mechanism, the bad guys find a way to get around it.
Device ID — the ability to uniquely identify and later recognize a user’s device — was one of the first tools enterprises used for authentication and fraud detection. Using regular and Adobe Flash cookies, you could tag a device and use that as the “something you have” component of the authentication process, thus replacing onerous hardware tokens. If a device was unknown, the enterprise could step up authentication measures.
Modern device ID solutions have become significantly more sophisticated than these early cookie-based solutions. They collect information on myriad device characteristics, both static and dynamic, including browser, operating system, internet connection and other properties. This allows security teams to create a unique fingerprint of the device, which can be used to authenticate customers or detect suspicious interactions.
While device ID remains an important and sometimes effective tool in the enterprise fraud detection arsenal, it is not nearly enough to constitute a complete fraud detection solution. Why is this?
Fraud Has Caught Up With Device ID Techniques
When device ID was first developed, bad actors quickly learned that they could copy cookies and use them on other devices, enabling them to appear legitimate. As the technique evolved to include things such as IP address and the type and version of browser and operating system, bad actors have reverse engineered device ID solutions and created increasingly detailed spoofing techniques to fool security algorithms.
Many malware strains today collect not only credentials, but also the data used to create a device ID. Bad actors can then manipulate their own device to appear to use the same browser extension, OS attributes and more to further impersonate their intended victim. This practice is known as device ID spoofing. Modern device ID solutions should include spoofing detection capabilities. Moreover, to keep up with the pace of sophisticated fraud activity, device ID spoofing detection must be updated daily based on ongoing research and threat intelligence.
RATs and Social Engineering
The eruption of remote access Trojans (RATs) and other similar threats has resulted in a new way for bad actors to avoid device ID-based fraud detection. An attacker using a RAT is actually using the victim’s device, which completely sidesteps any fraud detection capabilities based on device ID.
In addition to RATs, threat actors constantly develop schemes that take advantage of the weakest element of security strategy — humans — using social engineering tactics. Social engineering attacks such as business email compromise (BEC) target employees with access to company finances and trick them into making wire transfers to criminal bank accounts. In these cases, the fraudulent action comes from both the right device and the right user, something that a device ID-based fraud detection solution would be unable to detect.
Of course, the attacks that circumnavigate device ID-centric solutions are not yet simple enough to be conducted at scale. Fraudsters must invest significant time and research to complete these attacks successfully, but that doesn’t mean they should be overlooked. In fact, bad actors who employ these techniques generally target an institution’s highest-value accounts, making every successful attack potentially catastrophic.
Best Practices for Improving Fraud Strategies
What should an enterprise look for when implementing a fraud detection strategy? It should still include complex device ID as an integral feature, but it should be paired with a strong device ID spoofing tool that includes ongoing threat research and automatically adapts to new threats.
Perhaps more importantly, enterprises should think of device ID as just one tool in a multilayered identification toolbox. Device ID solutions should include additional indicators of fraudulent activity relative to the user, device, behavior or session. These can include behavioral biometrics, malware detection, phishing detection and global identity networks exposing repeated usage patterns over the multitude of these perspectives. It’s also important to consider ongoing transaction monitoring to identify accounts that might be compromised by social engineering.
From a wider security perspective, enterprises should always be wary of one-trick pony solutions. Any solution that uses device ID, biometrics or malware detection exclusively will never be enough to prevent fraud. Multilayered security solutions provide the depth needed to defeat the bad actors of today and tomorrow because they are infused with many layers of cognitive fraud detection and analytics to help prevent digital identity fraud.
In addition to highly complex device ID tools with spoofing detection, these solutions include ongoing global threat intelligence research, behavioral biometrics, malware detection, RAT detection and more. The security layers are pre-integrated, both on the technical level and on the derived risk balancing level, which helps organizations avoid the potential pitfalls of device ID-based fraud protection so they can offer their customers a seamless user experience.