skip to main content


Title: Impeding Behavior-based Malware Analysis via Replacement Attacks to Malware Specications
As the underground market of malware flourishes, there is an exponential increase in the number and diversity of malware. A crucial question in malware analysis research is how to define malware specifications or signatures that faithfully describe similar malicious intent and also clearly stand out from other programs. Although the traditional malware specifications based on syntactic signatures are efficient, they can be easily defeated by various obfuscation techniques. Since the malicious behavior is often stable across similar malware instances, behavior-based specifications which capture real malicious characteristics during run time, have become more prevalent in anti-malware tasks, such as malware detection and malware clustering. This kind of specification is typically extracted from the system call dependence graph that a malware sample invokes. In this paper, we present replacement attacks to cam- ouflage similar behaviors by poisoning behavior-based specifications. The key method of our attacks is to replace a system call dependence graph to its semantically equivalent variants so that the similar malware samples within one family turn out to be different. As a result, malware analysts have to put more efforts into reexamining the similar samples which may have been investigated before. We distil general attacking strategies by mining more than 5, 200 malware samples’ behavior specifications and implement a compiler-level prototype to automate replacement attacks. Experiments on 960 real malware samples demonstrate the effectiveness of our approach to impede various behavior-based mal- ware analysis tasks, such as similarity comparison and malware clustering. In the end, we also discuss possible countermeasures in order to strengthen existing mal- ware defense.  more » « less
Award ID(s):
1223710
NSF-PAR ID:
10066921
Author(s) / Creator(s):
; ; ; ; ;
Date Published:
Journal Name:
Journal of computer virology and hacking techniques
ISSN:
2263-8733
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Machine learning techniques are widely used in addition to signatures and heuristics to increase the detection rate of anti-malware software, as they automate the creation of detection models, making it possible to handle an ever-increasing number of new malware samples. In order to foil the analysis of anti-malware systems and evade detection, malware uses packing and other forms of obfuscation. However, few realize that benign applications use packing and obfuscation as well, to protect intellectual property and prevent license abuse. In this paper, we study how machine learning based on static analysis features operates on packed samples. Malware researchers have often assumed that packing would prevent machine learning techniques from building effective classifiers. However, both industry and academia have published results that show that machine-learning-based classifiers can achieve good detection rates, leading many experts to think that classifiers are simply detecting the fact that a sample is packed, as packing is more prevalent in malicious samples. We show that, different from what is commonly assumed, packers do preserve some information when packing programs that is “useful” for malware classification. However, this information does not necessarily capture the sample’s behavior. We demonstrate that the signals extracted from packed executables are not rich enough for machine-learning-based models to (1) generalize their knowledge to operate on unseen packers, and (2) be robust against adversarial examples. We also show that a na¨ıve application of machine learning techniques results in a substantial number of false positives, which, in turn, might have resulted 
    more » « less
  2. null (Ed.)
    Malicious software, popularly known as malware, is a serious threat to modern computing systems. A comprehensive cybercrime study by Ponemon Institute highlights that malware is the most expensive attack for organizations, with an average revenue loss of $2.6 million per organization in 2018 (11% increase compared to 2017). Recent high-profile malware attacks coupled with serious economic implications have dramatically changed our perception of threat from malware. Software-based solutions, such as anti-virus programs, are not effective since they rely on matching patterns (signatures) that can be easily fooled by carefully crafted malware with obfuscation or other deviation capabilities. Moreover, software-based solutions are not fast enough for real-time malware detection in safety-critical systems. In this paper, we investigate promising approaches for hardware-assisted malware detection using machine learning. Specifically, we explore how machine learning can be effective for malware detection utilizing hardware performance counters, embedded trace buffer as well as on-chip network traffic analysis. 
    more » « less
  3. Machine learning (ML) models have shown promise in classifying raw executable files (binaries) as malicious or benign with high accuracy. This has led to the increasing influence of ML-based classification methods in academic and real-world malware detection, a critical tool in cybersecurity. However, previous work provoked caution by creating variants of malicious binaries, referred to as adversarial examples, that are transformed in a functionality-preserving way to evade detection. In this work, we investigate the effectiveness of using adversarial training methods to create malware-classification models that are more robust to some state-of-the-art attacks. To train our most robust models, we significantly increase the efficiency and scale of creating adversarial examples to make adversarial training practical, which has not been done before in raw-binary malware detectors. We then analyze the effects of varying the length of adversarial training, as well as analyze the effects of training with various types of attacks. We find that data augmentation does not deter state-of-the-art attacks, but that using a generic gradient-guided method, used in other discrete domains, does improve robustness. We also show that in most cases, models can be made more robust to malware-domain attacks by adversarially training them with lower-effort versions of the same attack. In the best case, we reduce one state-of-the-art attack’s success rate from 90% to 5%. We also find that training with some types of attacks can increase robustness to other types of attacks. Finally, we discuss insights gained from our results, and how they can be used to more effectively train robust malware detectors. 
    more » « less
  4. Machine learning (ML) models have shown promise in classifying raw executable files (binaries) as malicious or benign with high accuracy. This has led to the increasing influence of ML-based classification methods in academic and real-world malware detection, a critical tool in cybersecurity. However, previous work provoked caution by creating variants of malicious binaries, referred to as adversarial examples, that are transformed in a functionality-preserving way to evade detection. In this work, we investigate the effectiveness of using adversarial training methods to create malware-classification models that are more robust to some state-of-the-art attacks. To train our most robust models, we significantly increase the efficiency and scale of creating adversarial examples to make adversarial training practical, which has not been done before in raw-binary malware detectors. We then analyze the effects of varying the length of adversarial training, as well as analyze the effects of training with various types of attacks. We find that data augmentation does not deter state-of-the-art attacks, but that using a generic gradient-guided method, used in other discrete domains, does improve robustness. We also show that in most cases, models can be made more robust to malware-domain attacks by adversarially training them with lower-effort versions of the same attack. In the best case, we reduce one state-of-the-art attack’s success rate from 90% to 5%. We also find that training with some types of attacks can increase robustness to other types of attacks. Finally, we discuss insights gained from our results, and how they can be used to more effectively train robust malware detectors. 
    more » « less
  5. null (Ed.)
    Android applications rely heavily on strings for sensitive operations like reflection, access to system resources, URL connections, database access, among others. Thus, insight into application behavior can be gained through not only an analysis of what strings an application creates but also the structure of the computation used to create theses strings, and in what manner are these strings used. In this paper we introduce a static analysis of Android applications to discover strings, how they are created, and their usage. The output of our static analysis contains all of this information in the form of a graph which we call a string computation. We leverage the results to classify individual application behavior with respect to malicious or benign intent. Unlike previous work that has focused only on extraction of string values, our approach leverages the structure of the computation used to generate string values as features to perform classification of Android applications. That is, we use none of the static analysis computed string values, rather using only the graph structures of created strings to do classification of an arbitrary Android application as malware or benign. Our results show that leveraging string computation structures as features can yield precision and recall rates as high as 97% on modern malware. We also provide baseline results against other malware detection tools and techniques to classify the same corpus of applications. 
    more » « less