Website defacement — the act of visibly altering the pages of a website, notably in the aftermath of a political event to advance the political agenda of a threat actor— has been explored in our various research works. We broke down top defacement campaigns in a previous paper and, in another post, emphasized how machine learning in our security research tool can help Computer Emergency Readiness Teams (CERTs)/Computer Security Incident Response Teams (CSIRTs) and web administrators prepare for such attacks. The latter took off from the analysis done in our most recent paper, Web Defacement Campaigns Uncovered: Gaining Insights From Deface Pages Using DefPloreX-NG. Here we expound on why machine learning (ML) was an ideal method for our analysis to better understand how web defacers operate and organize themselves.Read More
The history of antimalware security solutions has shown that malware detection is like a cat-and-mouse game. For every new detection technique, there’s a new evasion method. When signature detection was invented, cybercriminals used packers, compressors, metamorphism, polymorphism, and obfuscation to evade it. Meanwhile, API hooking and code injection methods were developed to evade behavior detection. By the time security solutions started using machine learning (ML)-based detection technologies, it was already expected that cybercriminals would develop new tricks to evade ML.
To be one step ahead of cybercriminals, one method of enhancing an ML system to counter evasion tactics is generating adversarial samples, which are input data modified to cause an ML system to incorrectly classify it. Interestingly, while adversarial samples can be designed to cause ML systems to malfunction, they can also, as a result, be used to improve the efficiency of ML systems.Read More
By employing machine learning algorithms, we were able to discover an enormous certificate signing abuse by BrowseFox, a potentially unwanted application (PUA) detected by Trend Micro as PUA_BROWSEFOX.SMC. BrowseFox is a marketing adware plugin that illicitly injects pop-up ads and discount deals. While it uses a legitimate software process, the adware plugin may be exploited…Read More
The ACM ASIA Conference on Computer and Communications Security (ACM ASIACCS) is an avenue for cybersecurity research breakthroughs, techniques, and tools. At the ACM ASIACCS 2018 in Incheon, South Korea, we presented our research using DefPloreX-NG, a tool for identifying and tracking web defacement campaigns using historical and live data. “DefPloreX-NG” is a play on the phrase “defacement explorer.” The appended “NG” acronym means “Next Generation,” signifying improvements from the previous version of the tool. DefPloreX-NG is equipped with an enhanced machine learning algorithm and new visualization templates to give security analysts and other professionals a better understanding of web defacement campaigns.Read More
Using a machine learning system, we analyzed 3 million software downloads, involving hundreds of thousands of internet-connected machines, and provide insights in this three-part blog series. In the first part of this series, we took a closer look at unpopular software downloads and the risks they pose to organizations. We also briefly mentioned the problem regarding code signing abuse, which we will elaborate on in this post.
Code signing is the practice of cryptographically signing software with the intent of giving the operating system (like Windows) an efficient and precise way to discriminate between a legitimate application (like an installer for Microsoft Office) and malicious software. All modern operating systems and browsers automatically verify signatures by means of the concept of a certificate chain.
Valid certificates are issued or signed by trusted certification authorities (CAs), which are backed up by parent CAs. This mechanism relies entirely and strictly on the concept of trust. We assume that malware operators are, by definition, untrustworthy entities. Supposedly, these untrustworthy entities have no access to valid certificates. However, our analysis shows that is not the case.Read More