Third part of the multipart “Defense against malware”

The workstations of our malware analysts do not differ from others in Hornetsecurity’s offices, even though the Security Lab is referred to as a “laboratory”. Erlenmeyer flasks, test tubes and Bunsen burners are not to be found, but quite normal computers. The work is done virtually, in sandboxes or by analyzing the data traffic. Nevertheless, the importance of malware analysts should not be underestimated, as it ensures that Hornetsecurity’s defense systems are always as up-to-date as possible and maintain the highest quality standard.

But what is the procedure for analyzing malware? Usually there is a very large, continuous stream of data to analyze. The main task is to extract valuable information from the raw data, process it and make it “intelligent”. To this end, analysts use various tools and programs to answer specific questions: What are the objectives of malware? Which characteristics are typical for the investigated malware? Is there any evidence of the attacker(s)? Ideally, actions can be derived from the findings such as writing new filter rules or creating algorithms.

Two different types of analysis

Two ways of analyzing malware are presented in more detail here. In static analysis, the code itself is viewed without executing the malware, while in dynamic analysis, the behavior of the malicious code is tracked in a secure environment.

In the static analysis, the analysts break down the malware to the smallest detail in order to draw conclusions from the code itself. For example, significant strings are extracted or shell scripts are started and further results are generated with disassemblers. Here you can find information on the activities of the malware and which features it shows, the so-called Indicators of Compromise (IoC). Based on the findings, the individual filter systems can be updated to prevent further attacks by this and similar malware as quickly as possible.

One possibility for dynamic analysis is to let the malicious code perform its task in the secure environment of a sandbox. This method can be well automated to obtain certain results. The filter systems can be updated based on these results. Does the code change certain files, does it make changes in the registry or has it generally adapted the system settings to DNS servers, for example? Who does the malware contact? These and other questions can be answered in the following way.

Hornetsecurity News

Stay in touch

Sign up to get the latest News about Cloud Security.

Various possibilities of use

The most obvious application of the data obtained from malware analysis for IT security companies is to improve their defense methods and thus better protect their customers from attacks. To do this, analysts extract certain binary patterns and use them to create so-called Yara rules with which malware samples can be found, categorized and grouped. Behavior signatures applied in the sandbox can detect and categorize certain behavior patterns of malicious code.

An example: In the sandbox, an Office document in the file attachment is opened. There the behavioral signatures recognize that the document to be examined begins to collect and send information about user accounts. If this analysis takes place in a cloud-based environment, it is then possible to intercept the conspicuous emails and thus completely block the attacks.

All of these and many other defense measures should help to intercept and prevent an attack at the earliest possible point so that the damage caused by malware is as small as possible or, better yet, does not occur at all.

Much of the raw data obtained by malware analysis and the findings derived from it are also useful for general prevention. Research projects can benefit from this and make their scientifically-sound results available to the general public. In addition, the publication of malware analyses also serves to educate the public. Increasing knowledge about the approaches of cyber attacks and malware attacks helps to limit their success rates.