Contributed to this research: Segev Fogel, Amir Gendler and Nethanella Messer.

IBM Trusteer researchers continually monitor the evolution and attack tactics in the banking sector. In a recent analysis, our team found that an Ursnif (aka Gozi) banking Trojan variant is being used in the wild to target online banking users in Italy with mobile malware. Aside from the Ursnif infection on the victim’s desktop, the malware tricks victims into fetching a mobile app from a fake Google Play page and infects their mobile device with the Cerberus Android malware.

The Cerberus malware component of the attack is used by Ursnif’s operators to receive two-factor authentication codes sent by banks to their users when account updates and money transfer transactions are being confirmed in real-time. Cerberus also possesses other features and can enable the attacker to obtain the lock-screen code and remotely control the device.

Cerberus is an overlay-type mobile malware that emerged in mid-2019 but initially lacked advanced capabilities. It has evolved over time to eventually feature the ability to hijack SMS content and control devices remotely, alongside other sophisticated data theft features. Cerberus was peddled in the underground as commodity malware until the summer of 2020, taking over the market share of Anubis, a previous pay-per-use malware.

In September 2020, Cerberus’ development team decided to disband, spurring an auction attempt that aimed to sell off the source code to the highest bidder, starting at $100,000. The code did not sell but was instead shared with the malware’s customer base, which meant it was publicly leaked. That intentional release of the source code gave rise to numerous malware campaigns involving Cerberus and likely also led to this combined attack with the Ursnif banking Trojan.

A Combination Attack From Desktop to Smartphone

Ursnif is a very long-standing staple in the cybercrime arena, possibly the oldest banking Trojan that’s still active today. Recent campaigns featuring this malware have been most notable in Italy, where it is typically delivered to business email recipients in attachments that purport to carry invoices, delivery notices or other business correspondence. The infection chain commonly involves poisoned macros, getting past email controls by featuring productivity files most organizations use. In some campaigns, the attackers keep access to the infection zone limited to Italian-based IP addresses only.

Once infected by the Ursnif malware and upon attempting to access their online banking account, victims are advised, via web injection, that they won’t be able to continue to use their bank’s services without downloading a security app. To obtain that app, they are shown a QR code and instructed to scan it with their phone’s camera.

Figure 1: Web injection instructing infected users to download a mobile app

Looking into the QR code provided through the injection, we found a Base64 encoded string with the details.

Figure 2: Malicious QR code’s content is a Base64-encoded string

If users scan the QR code, they will open a web page on their smartphone and be sent to a fake Google Play page featuring a corresponding banking app logo of the banking brand the victim originally attempted to access. The campaign, in this case, included a number of domains that were most likely registered for that purpose and reported in other malicious activity in the past, such as hxxps://[BANK BRAND].

Each of the domains hosting the fake Google Play pages used similar words or typo-squatting to appear legitimate. Some examples are:


These malicious domains have been flagged on VirusTotal for a few months, with more reports accumulating over time. Reports on the malicious Android Packages (APKs) that conceal the Cerberus malware spread in this campaign have been flagging it since at least late-2020.

In cases of users who do not successfully scan the QR code, they are asked to provide their telephone number and subsequently receive an SMS message with a download link to fetch the malicious application, which warns users about a potential service interruption if they fail to obtain the app.

Figure 3: Web injection instructing infected users to provide their phone number

In the background, the injection’s code couples the phone number inserted by the victim with the bot ID the Ursnif malware assigned to that infected desktop, the bank’s name the victim uses and their login credentials as grabbed by Ursnif.

Notice the use of the word ‘Jambo’ in parts of the code. It is most likely that Ursnif’s operators wrote a jQuery library to simplify HTML Document Object Model tree traversal and manipulation, using it to orchestrate their injections. Fraudsters can use the library to define the amounts to transfer from accounts and other parameters of the fraudulent transaction.

Figure 4: Injection code sending Cerberus infection URL to Ursnif-infected victims

If the victims submit their phone number on the web injection, the remote server will send back a download URL for them to unknowingly download the Cerberus malware. This injection also keeps the victims’ device identifiers linked to their bot ID and account credentials.

Figure 5: Injection code sending Cerberus infection URL to Ursnif-infected victims, additional view

Cerberus in Action

Cerberus campaigns have already been detected spreading through the official Google Play store in the past, but this distribution attempts to land on victim devices through a third-party source — the attacker’s domains. The option to sideload APKs is not enabled by default on Android devices, and the choice to deliver the malware from a non-official source may have limited the spread of the campaign to a larger number of devices.

When Cerberus is downloaded to a new device, it takes into account the original bank name the victim attempted to access when the infection process was initiated. A JavaScript function includes those details and ensures the victim continues to see a consistent message.

Here too, the ‘Jambo’ word repeats throughout the function, calling into action the jQuery library that orchestrates the malware’s script-based activity.

Figure 6: JavaScript function fetches Cerberus malware

Cerberus is being used here only as the component that allows the attackers to bypass the bank’s SMS-code verification challenge. The fraudulent transaction itself takes place on the victims’ infected desktops (Windows-based devices). While most fraud is in-session using Gozi SOCK proxy capability, some access to the victim’s account came from other devices.

Ursnif’s C2 Communications

The basics of Ursnif’s command and control (C2) communications are also carried out through the same channels. Jambo.getScript sends information to srv_dom, which is the malware’s injection server in this case, used to manage the man-in-the-browser activity.

Figure 7: Injection server communications

The core commands botmasters can launch come in where string ‘step=’ appears. Some of the available bot actions are:

Command Description
ADD_INFO Send data to C2: token, SMS content, telephone, download an application.
ASK Send communication to the C2.
GET_DROP Check account balance on the victim’s bank account.
GOOD_TRF Attempt to initiate a money transfer transaction.
LOGIN Send victim’s login information to attacker’s C2 server.
PING Check if the infected machine is currently online.
Scroll to view full table

IBAN Swapping Back in Style

On the infected desktop, we are back to seeing familiar activity from the Ursnif Trojan. Since it hooks the internet browser, it takes different steps to manipulate what victims see on their screens and have them click on elements that launch the Trojan’s resources into action.

One of the actions Ursnif wishes to take here is to automate transactions that start on the desktop’s browser. To do that, it is designed to swap the international bank account number (IBAN) and bank identifier code (BIC) numbers from legitimate transactions for an IBAN of an account the fraudster controls.

To launch its fraudulent transaction flow, Ursnif needs to start a function that would be clicked by the infected victim. It, therefore, attempts to replace a login button from the original bank’s webpage and plant its own button that the victim will click. The function launched is named ‘hookPay()’:

Figure 8: hookPay() function – Ursnif replaces IBAN number in legitimate transactions

The function being used to swap the IBAN and plan the transaction parameters is called ‘makeTrf()’. The amount being transferred is set to move forward if the account’s balance is higher than €3,000.

Figure 9: makeTrf() function – Ursnif sets up the fraudulent transaction’s parameters

Injections Adapt to Security Challenge

The configuration file in this campaign targeted the customers of banking institutions in Italy, specifically business banking services. On top of that, the attackers were after e-wallet and e-commerce credentials.

Web injections were adapted to each target’s security challenge; for example, an injection instructing victims to provide numbers from a hard token.

Figure 10: Adapting web injection social engineering to security challenge

Victims are asked to enter the code they received into the web injection and are given a 90-second time-lapse to do that, likely also adapted to the time allotted by the targeted bank or service provider:

Figure 11: Adapting web injection social engineering to security challenge, additional view

After receiving the data from the victim, the malware sends data to the C2 server, including authorization token, SMS content, telephone number and account login information. It then shows a .gif file that makes it appear as if the web browser is loading something. After a couple of seconds, the .gif file is hidden, and the malware continues the login process in the background.

To prevent victims from accessing the account and discovering the fraudulent activity before it is finalized, Ursnif presents a maintenance notice on the account. This notice can effectively prevent the victim from accessing the account from the infected device.

Figure 12: Victims are denied access to their bank account to hide fraudulent activity

Something Old, Something New — The Ursnif-Cerberus Combo

Banking Trojan operators have always been fans of fraud they can automate. The rollout of two-factor authentication and strong transaction authorization schemes by online banking services across the globe have caused this entire threat actor class to rethink their tactics, techniques and procedures. Over time, the incorporation of mobile malware into the overall scheme of banking Trojan fraud has become a must, since it is the only way to complete transactions. The hindrance remains that malware operators have to continue to find ways to infect more mobile devices, especially when getting into official app stores has been getting harder. Also, activating the victim for the initial setup of the automation process is another place where the criminal can fail. Fortunately, these are also the places where defenders can help prevent fraud.

Seeing Ursnif using Cerberus as its mobile malware component is new, but it is not surprising in the banking Trojan arena. Banking Trojan operators are constantly shifting tactics, but the strategy remains the same — they have to gain access to victims’ smartphones if they hope to get through security controls applied to banking and other services consumed online. Using Cerberus is also expected since the code was leaked and gave the option to any malware operator to make use of it against unsuspecting victims.

IBM Security Trusteer helps organizations detect fraud, authenticate users and establish identity trust across the omnichannel customer journey. Through cloud-based intelligence, backed by artificial intelligence and patented machine learning, Trusteer provides a holistic approach to identifying new and existing customers, while improving the user experience. More than 500 leading organizations rely on Trusteer to help secure their customers’ digital journey and support business growth. To learn more visit:

To keep malware off your mobile devices, follow some security hygiene basics:

  • Don’t jailbreak a smartphone
  • Only download apps from Google Play’s official store
  • If you download from a URL, get your bank’s application via your bank’s website
  • Don’t enable sideloading; your bank or service provider will not ask you to load applications from unofficial sources
  • Check who is the developer of the app you are downloading; if it does not look right, abort the download
  • Be wary of excessive app permissions: Only allow apps to use your device for the purpose you require and not for unrelated activities
  • If it looks like there’s a new security requirement from your bank, close the browser window and call your bank with the number on the back of your card to verify what’s needed
  • If a transaction you attempted to carry out is stopped by an apparent ‘maintenance’ issue, attempt to access the account from a different device or call your bank.


C2 Servers





MD5 Gozi: b6921ce0f1b94a938acb6896cc8daeba
MD5 Cerberus + APK:

Phishing domains and C&C servers:



IP addresses:

SOCKS Proxy:


More from Malware

Ongoing ITG05 operations leverage evolving malware arsenal in global campaigns

13 min read - As of March 2024, X-Force is tracking multiple ongoing ITG05 phishing campaigns featuring lure documents crafted to imitate authentic documents of government and non-governmental organizations (NGOs) in Europe, the South Caucasus, Central Asia, and North and South America. The uncovered lures include a mixture of internal and publicly available documents, as well as possible actor-generated documents associated with finance, critical infrastructure, executive engagements, cyber security, maritime security, healthcare, business, and defense industrial production. Beginning in November 2023, X-Force observed ITG05…

X-Force Threat Intelligence Index 2024 reveals stolen credentials as top risk, with AI attacks on the horizon

4 min read - Every year, IBM X-Force analysts assess the data collected across all our security disciplines to create the IBM X-Force Threat Intelligence Index, our annual report that plots changes in the cyber threat landscape to reveal trends and help clients proactively put security measures in place. Among the many noteworthy findings in the 2024 edition of the X-Force report, three major trends stand out that we’re advising security professionals and CISOs to observe: A sharp increase in abuse of valid accounts…

Hive0051’s large scale malicious operations enabled by synchronized multi-channel DNS fluxing

12 min read - For the last year and a half, IBM X-Force has actively monitored the evolution of Hive0051’s malware capabilities. This Russian threat actor has accelerated its development efforts to support expanding operations since the onset of the Ukraine conflict. Recent analysis identified three key changes to capabilities: an improved multi-channel approach to DNS fluxing, obfuscated multi-stage scripts, and the use of fileless PowerShell variants of the Gamma malware. As of October 2023, IBM X-Force has also observed a significant increase in…

Topic updates

Get email updates and stay ahead of the latest threats to the security landscape, thought leadership and research.
Subscribe today