Ursnif is a banking Trojan that usually infects corporate victims. It is based on an old malware but was substantially updated over the years and became quite powerful. Today Ursnif is one of the most widely spread banking Trojans in the world.

First seen
1 January, 2014
Last seen
4 October, 2022
Also known as
Global rank
Week rank
Month rank

What is Ursnif malware?

Ursnif, also known as Gozi, is one of the most widely spread banking trojans – it is aimed at stealing banking credentials and usually targets corporate victims. The malware was developed based on the leaked source code of a fairly old Gozi-ISFB trojan.

The trojan was registered for the first time in 2014 when the Gozi-ISFB code got leaked. Since then, Ursnif has been evolving and becoming more powerful, which lead it to become one of the top used banking trojans today.

General description of Ursnif trojan

Ursnif Trojan is a dangerous malware that can collect the system activity of the victims, record keystrokes, and keep track of network traffic and browser activity. The malware stores the data in an archive before sending it to the C2.

The malware uses malicious Microsoft Office documents to get into the users’ machine and requires macros to be activated. Once opened, the document will prompt the user to enable macros. If the user plays along with the instruction, the malware drops a VB script into the temp directory of the current user, upon which it is automatically decoded, and the malicious payload is downloaded.

According to the analysis, some versions of Ursnif contain a macro that is programmed to check the country using the Application. International MS Office property. If the result does not correspond to a list of pre-selected countries, the malware terminates its execution.

Interestingly, the malware terminates execution if it detects that it’s being launched on a virtual machine. Hackers implement this precaution technique in order to complicate the analysis process and, hopefully, prevent the effective development of countermeasures.

Ursnif malware analysis

A video is available at ANY.RUN malware analysis service allows us to see a simulation of the malware execution in a lot of detail. YOu can also investigate other malware like Hawkeye or Raccoon.

ursnif gozi dreambot execution graph Figure 1: A visual process graph generated by ANY.RUN shows the lifecycle of Urnsnif

How to avoid infection by Ursnif?

The best way to stay safe from Ursnif is to keep the macros turned off and not turn them on if prompted by a Microsoft Office file downloaded from an untrustworthy source, such as an email from the unknown sender. In addition, following good techniques of staying safe online such as not downloading files from suspicious emails, is another great way to avoid infection.

Ursnif execution process

In the case of our simulation, the execution of the malware starts when the user opens a Word or Excel file and enables the macro. Ursnif uses the browser's COM object to connect to its C2 server and receive additional data.

Based on the analysis, Ursnif trojan uses exploits to start legitimate software like Outlook, which in turn launches cmd.exe only to spawn a PowerShell script. If a strike is directed at select countries, the malware checks where the victim is from during this stage. Then, the PowerShell script downloads and executes the final payload, which is Ursnif itself. Lastly, the loader starts malicious activities and injects its code into the explorer.exe process.

After installation, the malware will try to inject into an active explorer.exe process to establish persistence. If the injection fails, Ursnif will launch a new svchost.exe process and will inject itself instead. this technique appears to be a useful pointer for detection. After that, Ursnif will hook the APIs of common web browsers such as Chrome, Opera, Internet Explorer, and Firefox. The loader uses the browsers' COM object to communicate to its C2 server. Then, the malware will begin monitoring web activity and steal the payment information as soon as the victim visits a banking or a payment webpage. Then Ursnif sends collected data to a C2 server via the IE COM object.

Communication with C&C

In order to prevent domain name disclosure, the malware generates the domain names locally using the technique of the Domain Generation Algorithm (DGA) instead of them being hardcoded. Uniquely, the malware gathers information for domain name generation in the DGA process by taking bits of text from popular websites. If you decrypt the URL in the script, you may get the data sent to the C2 server.

The malware is also known to be able to execute commands received from the control server.

How to detect Ursnif using ANY.RUN?

Ursnif uses COM objects to execute the malware's payload, and usually, it runs multiple iexplorer.exe processes. The loader creates a COM object that is a hidden API function. Knowing this information, take a look at the process tree after a while during execution, and determine either sample is Ursnif or not. Check the script to find out if a suspicious URL corresponds to malware activity.

ursnif process tree Figure 2: Ursnif process tree


Based on the source code of another malware that is already almost a decade old, Ursnif is a prime example of the fact that “old” does not mean ineffective when it comes to trojans.

On the contrary, despite its age, this malware is capable of launching devastating cyber attacks and managed to become one of the most popular banking trojans in the world. In addition to its powerful trojan functionality, the loader takes active actions to prevent researchers from studying it. Thankfully, malware hunting services like ANY.RUN allows researchers to study this malware in-depth and respond with appropriate countermeasures.


IP addresses


Adwind screenshot
adwind trojan
Adwind RAT, sometimes also called Unrecom, Sockrat, Frutas, jRat, and JSocket, is a Malware As A Service Remote Access Trojan that attackers can use to collect information from infected machines. It was one of the most popular RATs in the market in 2015.
Read More
Agent Tesla screenshot
Agent Tesla
agenttesla trojan rat stealer
Agent Tesla is spyware that collects information about the actions of its victims by recording keystrokes and user interactions. It is falsely marketed as a legitimate software on the dedicated website where this malware is sold.
Read More
Ave Maria screenshot
Ave Maria
avemaria stealer trojan rat
Ave Maria malware is a Remote Access Trojan that is also called WARZONE RAT. Hackers use it to control the PCs of their victims remotely and steal information from infected PCs. For example, they can remotely activate the camera to take pictures of a victim and send them to a control server.
Read More
Azorult screenshot
azorult trojan rat
AZORult can steal banking information, including passwords and credit card details, as well as cryptocurrency. This constantly updated information stealer malware should not be taken lightly, as it continues to be an active threat.
Read More
Crimson RAT screenshot
Crimson RAT
crimson rat trojan
Crimson is a Remote Access Trojan — a malware that is used to take remote control of infected systems and steal data. This particular RAT is known to be used by a Pakistani founded cybergang that targets Indian military objects to steal sensitive information.
Read More
Danabot screenshot
danabot trojan stealer
Danabot is an advanced banking Trojan malware that was designed to steal financial information from victims. Out of the Trojans in the wild, this is one of the most advanced thanks to the modular design and a complex delivery method.
Read More