Traditional antivirus software relies on signatures to uniquely identify malicious files. Malware writers, on the other hand, have responded by developing obfuscation techniques with the goal of evading content-based detection. A consequence of this arms race is that numerous new malware instances are generated every day, thus limiting the effectiveness of static detection approaches. For effective and timely malware detection, signature-based mechanisms must be augmented with detection approaches that are harder to evade.
We introduce a novel detector that uses the information gathered by IBM’s QRadar SIEM (Security Information and Event Management) system and leverages anti-virus reports for automatically generating a labelled training set for identifying malware. Using this training set, our detector is able to automatically detect complex and dynamic patterns of suspicious machine behavior and issue high-quality security alerts. We believe that our approach can be used for providing a detection scheme that complements signature-based detection and is harder to circumvent.