Banking Trojans and the gangs that operate them continue to plague banks, individuals and organizations with fraudulent transactions facilitated by malware and social engineering schemes. At last check, cybercrime cost the global economy more than $600 billion in 2017 , and forecasts for 2018 predicted $1.5 trillion in losses.
No matter how you turn these numbers, they are a burden that keeps growing and encouraging a rife, complex industry of online crime.
Going Behind the Numbers of the ‘IBM X-Force Threat Intelligence Index’
Every year, increasingly organized cybercrime gangs shuffle their tactics, techniques and procedures (TTPs) to evade security controls on the micro level and law enforcement on the macro level. Behind each malware named on the top 10 chart below, codes are distributed and operated differently and focus on different parts of the globe. The chart is populated by organized cybercrime gangs that have ties to yet other cybercrime gangs, each doing its part to feed the perpetual supply chain of a digital financial crime economy.
In cybercrime, it can be said that the more things change, the more things stay the same. In 2018, however, I must admit I was finally surprised when two malware gangs that did not appear connected at first began openly collaborating. It thus became clearer than ever that the banking Trojan arena is dominated by groups from the same part of the world, by people who know each other and collaborate to orchestrate high-volume wire fraud.
To learn more about the malware that shaped 2018, let’s begin by looking at the top constituents of the gang-owned Trojan chart and drill down on information gathered by IBM Security for the top three.
Figure 1: Top 10 chart of the most active banking Trojan families in 2018 (source: IBM X-Force)
TrickBot, a banking Trojan operated by a Russia-based threat group, was one of the most aggressive Trojans of 2018. It targets banks across the globe with URL-heavy configurations that often include a large number of targeted bank brands from across the globe.
TrickBot’s operators focus on business banking and high-value accounts that are held with private banking and wealth management firms, but they also diversified in 2018 to include various e-commerce and cryptocurrency exchange platforms on their target lists.
According to IBM X-Force data that was gathered since TrickBot’s rise, no other financial Trojan is as consistently active in terms of infection campaigns and deployment of redirection attacks, indicating that its operators have ample resources and connections to develop and operate the malware in different parts of the world. Despite this overall capability, X-Force saw TrickBot sharpening its focus in 2018 and targeting a handful of countries in each campaign, keeping major economies such as the U.K. and the U.S. on almost every target list.
Intergang Collaboration With IcedID
Some of the trends in TrickBot’s activity in 2018 included collaboration with another banking Trojan, IcedID, which IBM X-Force discovered in September 2017, as well as operating the Ryuk ransomware, a subset of TrickBot’s botnet monetization strategy. These highlight a larger trend of intergang collaboration among Trojan operators striving to generate larger profits in spite of growing security control sophistication.
At first, TrickBot and IcedID appeared unrelated. But about eight months into IcedID’s existence, signs of a link between the two became apparent. In May 2018, X-Force researchers observed TrickBot dropping IcedID, whereas it had previously been dropped primarily by the Emotet Trojan, the same distributor that also drops TrickBot in different campaigns.
By August 2018, our researchers noted that IcedID had been upgraded to behave in a similar way to the TrickBot Trojan in terms of its deployment. The binary file was modified to become smaller and no longer featured embedded modules. The malware’s plugins were being fetched and loaded on demand after the Trojan was installed on infected devices. These changes made IcedID stealthier, modular and more similar to TrickBot.
In addition to its increased stealth level, IcedID also started encrypting its binary file content by obfuscating file names associated with its deployment on the endpoint. Also similar to TrickBot is IcedID’s event objects, which coordinate multiple threads of execution in Windows-based operating systems. IcedID began using named events to synchronize the execution between its core binary and the plugins selected for loading. When a plugin was called upon, it was fetched by its ID number from the attacker’s server and, when loaded, assigned a unique ID.
Although malware authors do sometimes copy from one another, our research indicates these modifications were not coincidental. Even if we only looked at the fact that TrickBot and IcedID fetch one another into infected devices, that would be indication enough that these Trojans are operated by teams that work together.
Ties between TrickBot and IcedID may have started years ago in a collaboration designed to help both groups maximize their illicit operations and profits. During the six-year activity phase of the Neverquest (aka Catch or Vawtrak) Trojan, it collaborated with the Dyre group to deliver Dyre malware to devices already infected with Neverquest.
The original Dyre group partly disbanded in late 2015, followed by the rise of TrickBot, which is believed to be the successor to Dyre. Neverquest halted operations following the arrest of one of its key members in 2016, after which the IcedID Trojan appeared. With the two featuring advanced capabilities and evident cybercrime connections with other gangs, their current-day collaboration likely started years ago.
The TrickBot-Ryuk Connection
Another TrickBot trend that started in 2018 is a connection with ransomware. Reminiscent of the Dridex Trojan’s links to the Locky and then BitPaymer ransomware, TrickBot began dropping ransomware called Ryuk. Unlike wide-cast nets that spread ransomware to as many email recipients as possible, Ryuk, like BitPaymer, is spread in targeted campaigns where attackers go through the typical advanced persistent threat (APT) kill chain and manually breach the network.
Ryuk attackers often go through reconnaissance stages, looking for valuable data to hijack. The goal: Infect established organizations with Ryuk and then demand large sums in ransom payments that average hundreds of thousands of dollars each.
Figure 2: Ryuk campaigns — a four-step routine to drop three different Trojans to target devices (source: IBM X-Force)
Upon investigating Ryuk’s code, it quickly became apparent that this ransomware was not entirely new. Ryuk closely resembles the Hermes ransomware that was linked with malicious activity by a nation-state-sponsored group called Lazarus (aka Hidden Cobra).
Is Ryuk connected to Hermes? That’s one possibility. It could also be that some Lazarus members collaborate with banking Trojan operators through cross-border partnerships to steal and launder large amounts of cybercrime money via Eastern Europe and Asia, or that someone with access to the Hermes code reused it to create Ryuk.
Whatever the source of Ryuk, it shows that TrickBot’s operators are diversifying their nefarious activity, continuing to focus heavily on the business sector and launching targeted attacks that press organizations to pay.
Figure 3: Collaboration between major malware gangs (source: IBM X-Force)
TrickBot TTPs and Evolution
In terms of its TTPs, TrickBot’s operators focus their efforts on businesses and, therefore, opt for distribution through booby-trapped productivity files and fake bank websites. After infection, TrickBot modules allow it to spread laterally in compromised networks and infect additional users.
TrickBot continues to use both server-side injections deployed on the fly from its attack server and redirection attacks hosted on its servers to hijack users and present them with a fake replica of their bank’s website.
In 2018, TrickBot’s developers added three new functions to the malware, facilitating the theft of Remote Desktop Protocol (RDP) credentials, Virtual Network Computing (VNC) credentials and PuTTY open-source terminal emulator credentials. It steals Windows system reliability and performance data and features a new mechanism for storing its configuration and elliptic curve cryptography (ECC) public keys.
The TrickBot botnet is supported by what’s considered a mature infrastructure, where some campaigns featured 2,458 unique command-and-control (C&C) IP addresses used in 493 main configuration releases across 276 versions — all in one week.
X-Force expects to see TrickBot maintain its position on the global malware chart unless it is interrupted by law enforcement in 2019.
Gozi (aka Ursnif) has been highly active in the wild for more than a decade now, a rare occurrence in the cybercrime arena. The malware was first discovered in 2007, when it was operated by a closed group of developers and cybercriminals. At the time, it was used to target online banking users mostly in English-speaking countries.
Throughout the years, Gozi has gone through almost every phase a banking Trojan can go through. Its code was leaked in 2010, giving rise to other Trojans, such as Neverquest, that also dominated the cybercrime charts for years after. It was used in the Gozi-Prinimalka ordeal in 2012 and, in 2013, was fitted with a master boot record (MBR) rootkit to create high persistence through a computer’s MBR.
In 2016, X-Force reported about the rise of the GozNym hybrid, a two-headed beast spawned from the Nymaim malware and embedded with the Gozi financial fraud module. Starting in 2017, X-Force researchers reported that a new variation of Gozi was being tested in Australia: Gozi v3. The malware was based on the same code of the original Gozi ISFB but featured some modifications on the code injection level and attack tactics.
In 2018, Gozi v2 was the second-most active Trojan in the wild, working across the globe and in Japan. V2 is operated separately from the v3 version that continues to target banks in the Australia-New Zealand region. The malware is operated in a cybercrime-as-a-service model that allows different cybercriminals to use the botnet to conduct fraud.
To reach new victims, Gozi is distributed in document and spreadsheet attachments that prompt the user to enable macros. In recent campaigns, when the user complies, the macro runs the WMI Provider Host process (wmiprvse) to execute a malicious PowerShell script. The script is designed to fetch the payload and uses string concatenation to evade detection.
Recently, in the case of attack schemes against banks in Europe, Gozi delivered custom-tailored client-side code for each targeted bank brand users accessed, likening its tactics to redirection attacks in which each brand is targeted in a specific way.
Gozi’s distributors use malicious websites to host their resources but check the target device’s Geo IP to reduce the potential of exposure. If either Russian or Chinese keyboard settings are detected during its installation, the deployment ends.
This malware has been part of the top-most constituents of the global malware chart for the past five years, and X-Force expects to see this longtime staple of the organized cybercrime arena maintain its position on the chart in 2019.
Ramnit is a prolific banking Trojan that has been active in the wild since 2010. Ramnit started out as a self-replicating worm, leveraging removable drives and network shares to spread to new endpoints. As the project evolved, Ramnit morphed into a modular banking Trojan and started spreading via popular exploit kits such as Angler and RIG.
Although it was one of the most prominent Trojans between 2011 and 2014, Ramnit was targeted by law enforcement in 2015. While it was one of the only botnets to ever survive a coordinated disruption, its operators have not returned to the same level of activity since. In recent years, Ramnit has been an on-again-off-again operation, seeing long lulls in its cybercrime activity and narrowing its attack turf over time to focus mostly on the U.K., Canada and Japan.
2018: Reemergence and Intergang Collaboration
In 2018, the Ramnit Trojan returned to the cybercrime arena with revamped code and a new partner, a proxy malware known as Ngioweb. Ramnit’s developer modified its financial module to enhance its capabilities and changed the internal module’s name from “Demetra” to “Camellia.”
Ramnit’s 2018 comeback resulted in a reported infection of more than 100,000 devices within the span of two months, as part of an operation code-named “Black.” In this campaign, Ramnit went back to its worm roots and was used as a first-stage infection in a kill chain designed to amass a large proxy botnet for Ngioweb.
How good was this new partnership for Ramnit? We can only assume that it was used to create a massive proxy botnet that would resemble the Gameover Zeus botnet in its architecture. The Black campaign was short-lived, and by the end of 2018, Ramnit was linked with Emotet, Dridex and BitPaymer for using the same dropper as those Trojans and being used itself as a dropper for Dridex.
Configuration and code comments show that Ramnit is probably being developed by new team members. Configuration injects were modified to Lua programming and, in many cases, came bugged or unsophisticated. This was not the case for this malware in past years.
For its deployment routine, Ramnit began leveraging code that relies on PowerShell scripts in what’s known as reflective PE injection. Its modules are not pulled from a remote server but come packed with the core malware, and its reliance on a domain generation algorithm (DGA) has been modified to include hardcoded domains.
Will we continue to see Ramnit in 2019? X-Force researchers expect to see the same activity pattern for this malware with its come-and-go nature in Japan and Europe. Ramnit will likely drop from its current rank on the global Trojan chart and be overtaken by IcedID and newcomers like BackSwap and DanaBot.
Threat Landscape Staples
Banking Trojans have been a burdensome part of the cybercrime threat landscape for more than a decade now. The past five years have shown us that this breed of attackers is only becoming more sophisticated over time, incorporating technical knowledge with advance social engineering to focus schemes on victims that can yield the biggest profits: businesses, cryptocurrency and high-value individuals.
While previous years saw gangs operate as adversaries, occupy different turfs and even attack each other’s malware, our research from 2018 connected the major cybercrime gangs together in explicit collaboration. This trend is a negative sign that highlights how botnet operators join forces, revealing the resilience factor in these nefarious operations.
While it can be hard to detect this type of evolving malware, it’s possible to stop banking Trojans before they make it into your device or your organization. Proper security controls and user education, as well as planned incident response, can help keep this threat at bay and contain its detrimental effects if ever an account is taken over and robbed by highly experienced criminals.
To learn more about the top security threats of 2018 and what 2019 may have in store, download the “IBM X-Force Threat Intelligence Index.” Check out page 30 in the report for our expert team’s tips on mitigating threats and increasing preparedness for a possible breach.
Read the full “IBM X-Force Threat Intelligence Index”