Detection highlights · September 2026

FlowerStorm phishing kit, ETW tampering and 70+ new YARA rules

Izabela Komorowska Technical writer

· 7 min read

TL;DR

11New VTIs

70+New YARA rules

  • Five of the new VTIs follow the FlowerStorm phishing kit from CAPTCHA to phishing page.
  • New detections for ETW tampering, Linux dynamic linker hijacking, malicious Chrome extensions and Microsoft device-enrollment links.
  • Two phishing heuristics: abuse of public web services, and forms that require identification.

The Labs team at VMRay actively gathers publicly available data to identify any noteworthy malware developments that demand immediate attention. We complement this effort with our internal tracking and monitor events the security community reports to stay up-to-date with the latest changes in the cyber threat landscape.

In September 2026, the VMRay Labs team has been focused on the following areas:

Now, let’s delve into each topic for a more comprehensive understanding.

New VTIs

In a series of these blog posts, we introduced you to the concept of the VMRay Threat Identifiers (VTIs). In short, VTIs identify threatening or unusual behavior of the analyzed sample and rate the maliciousness on a scale of 1 to 5, with 5 being the most malicious. The VTI score, which greatly contributes to the ultimate Verdict of the sample, is presented to you in the VMRay Platform after a completed analysis. Here's a recap of the new VTIs that we added, or improved in the past month.

Detecting attempts to disable Windows Event Tracing providers

Category: Defense Evasion

MITRE ATT&CK® Technique: T1562.006

Event Tracing for Windows (ETW) records information about activity within the Windows operating system and its applications. ETW providers generate events for specific types of activity, such as processes starting or stopping, files or registry keys being accessed, authentication attempts, and application errors. Monitoring and security tools, including malware sandboxes, can collect and analyze these events to understand what is happening on a system.

Attackers may attempt to disable an ETW provider or alter its configuration to prevent certain events from being recorded. By interrupting this information flow, malware can make its activity less visible to security and monitoring tools and potentially evade detection.

ETW provider settings can be modified through several mechanisms, including PowerShell, Windows Management Instrumentation (WMI), and changes to the Windows Registry. Although these mechanisms also have legitimate administrative uses, an unexpected attempt to disable a provider may indicate defense-evasion activity. Our new VTI will now flag any process that attempts to disable Event Tracing for Windows.

Detecting the FlowerStorm phishing kit

FlowerStorm is a popular phishing kit with distinctive network communication patterns that can help identify its activity. After dissecting the kit and examining how its attack chain works, our Labs team developed five VTIs covering different stages of its behavior.

CAPTCHA-related behavior

Before displaying the phishing page, FlowerStorm presents the visitor with a CAPTCHA-style verification step. Two new VTIs detect communication associated with this stage:

  • Detecting submission of CAPTCHA verification data to a FlowerStorm server
  • Detecting retrieval of script links from a FlowerStorm server; these scripts are used to build and display the phishing page

Phishing page behavior

Two additional VTIs detect the subsequent loading and visitor-validation stages:

  • Detecting downloads of obfuscated JavaScript containing a FlowerStorm phishing page
  • Detecting visitor validation performed by a FlowerStorm server

Finally, a fifth, higher-level VTI correlates these individual behaviors. When the relevant stages are observed together, it identifies the overall activity as characteristic of the FlowerStorm phishing kit.

Figure 1

Four VTIs each detect one stage of FlowerStorm. A fifth recognizes the kit when the stages appear together.

CAPTCHA-related behavior
CAPTCHA data submittedVerification data sent to a FlowerStorm server
Script links retrievedScripts that build and display the phishing page
Phishing page behavior
Obfuscated JavaScriptDownloaded, containing the phishing page
Visitor validationPerformed by a FlowerStorm server
FlowerStorm phishing kitThe fifth VTI correlates the stages and identifies the kit

Detecting Dynamic Linker Hijacking on Linux

Category: Execution

MITRE ATT&CK® Technique: T1574.006

Linux applications do not always contain all the code they need to operate. Instead, they reuse code stored in shared libraries (.so files). When an application starts, the dynamic linker finds the required libraries, loads them into memory, and makes their functions available to the application.

Attackers can hijack this process by changing where the dynamic linker looks or which libraries it loads first. This can trick a legitimate application into loading malicious code instead of, or alongside, the expected library. Attackers may achieve this by modifying:

  • LD_PRELOAD, which instructs Linux to load a specified library before others,
  • /etc/ld.so.preload, which defines libraries to be loaded across the system,
  • LD_LIBRARY_PATH, which controls where the system searches for libraries.

Figure 2

One changed setting decides which library the linker loads, and the attacker’s code runs inside a legitimate process.

Attacker modifies
LD_PRELOAD/etc/ld.so.preloadLD_LIBRARY_PATH
Dynamic linkerFinds and loads libraries when an application starts
Malicious libraryLoaded instead of, or alongside, the expected one
Legitimate processRuns the malicious code, with the same permissions

The new VTI detects when a process attempts to hijack the dynamic linker through one of these methods.

If successful, the malicious code runs as part of a legitimate process and may gain the same permissions. Depending on the method used, this can help an attacker maintain persistence, conceal malicious activity, or gain elevated privileges. Our new VTI detects when a process attempts to hijack the Linux dynamic linker through one of several known methods.

Detecting possible abuse of public web services

Category: Heuristics

Attackers sometimes misuse legitimate public services, such as online form or page-building platforms, to support phishing campaigns. Because the page is hosted on a well-known and trusted domain, its link may appear harmless at first glance. This does not mean that the service itself has been compromised; rather, the attacker is using it for a purpose it was not designed to serve.

These pages often contain very little meaningful content and primarily direct visitors to another website, such as a phishing page. In this way, the trusted service acts as an intermediate step between the original link and the attacker’s destination.

We added a new VTI that identifies pages hosted on known public services when their content does not match the service’s expected use. The VTI looks for characteristics such as limited page content and a prominent link leading visitors elsewhere.

Detecting page containing forms that require identification

Category: Heuristics

Forms that ask visitors to identify themselves, for example, by entering an email address, username, or other personal information are common on legitimate websites. However, they are also frequently used on phishing pages to evade automated analysis, as visitors must enter and submit a valid email address before the subsequent content is displayed.

We expanded our coverage with a new informative VTI that detects when a page contains a form requiring identifying information, regardless of whether the page can be associated with a particular brand. This complements our existing detection for branded forms and provides broader visibility into potentially relevant form activity.

Because the presence of such a form is not inherently malicious, the VTI serves as a behavioral indicator. It helps analysts understand what the page is asking the visitor to do and can be combined with other suspicious signals.

Detecting loading/installation of Google Chrome extension

Category: Browser

MITRE ATT&CK® Technique: T1176

Browser extensions are add-ons that introduce new features or change how a browser behaves. While most extensions are legitimate, attackers can create malicious ones that redirect searches, display unwanted advertisements, monitor browsing activity, or modify the content of visited pages.

For example, the ChromeLoader malware family is known to use this capability. Instead of installing its extension through the Chrome Web Store, it starts Chrome with the command-line option. This tells the browser to load extension files directly from a specified folder on the system. An extension loaded this way is active for that browser session. To load it again, the malware can repeatedly launch Chrome with the same option, allowing the malicious extension to continue operating.

Our new VTI detects when a process tries to load a Chrome extension directly from files stored on the system. Although developers may use this option for legitimate testing, its unexpected use, particularly when Chrome is launched by an unusual process, can indicate malicious activity.

To support further investigation, the VTI provides IOCs such as:

  • The process that attempted to launch Chrome;
  • The path and filename of the extension;
  • The extension file as an artifact, when available; and
  • The associated Chrome process for context.

Detecting Microsoft device-enrollment links

Organizations use device enrollment to connect Windows computers to a management service. Once enrolled, a device can receive company applications, security settings, and other policies. Windows supports special links that can open this enrollment process directly from a webpage.

Although these links are commonly used by IT departments, attackers may also place them on deceptive pages to direct a device toward an unauthorized management server. If successful, this could allow the attacker to apply unwanted settings or otherwise interfere with how the device is managed.

Our new VTI detects when a webpage contains a Microsoft device-enrollment link. Because such links can also serve legitimate purposes, their presence alone is treated as a suspicious signal. To identify higher-risk cases, the Platform correlates this behavior with additional warning signs, including recently registered or lookalike domains, domain endings commonly associated with phishing, and reputation signals for URLs observed in the page’s network traffic or embedded content.

VMRay Threat Identifiers table in the VMRay Platform for a page with a Microsoft device-enrollment link: the 2/5 heuristic “Page contains Microsoft enrollment link” sits alongside a 5/5 combined detection that the page is a phishing website.
The new VTI in the VMRay Platform. On its own it scores 2/5; combined with other detections, the page is rated 5/5 as a phishing website.

YARA Rules Update

Our hunt for new, undetected malware samples never stops. In the past month, we added many fresh YARA rules to strengthen detection across a wide range of threats. This month, we’re continuing that momentum with 70+ new rules, focused on delivering a solid drop of high-quality detections. Here’s a quick preview of what we're shipping this month.

New YARA detections for:

Spyware

  • AokigaharaStealer
  • AsterStealer
  • BlackStealer
  • DeekStealer
  • f4maxStealer
  • GoTelegramStealer
  • PostureStealer
  • RobHookStealer
  • SBStealer
  • ShiverStealer
  • WaterStealer
  • YoucefGrabber
  • Zlogger
  • Zer0dayStealer
  • AestheticStealer
  • AkhassatStealer
  • FearStealer
  • KynixStealer
  • OverdriveStealer
  • OxideHarvest
  • QvoidStealer
  • RLStealer
  • XillenStealer
  • ZyxStealer
  • CtfCookieExtractor
  • DraculaStealer
  • LLHNStealer
  • NimdesStealer
  • RevStealer
  • BurayaStealer
  • ChayLaiStealer
  • EnhancedStealer
  • NWHStealer
  • NeedleStealer
  • HypeAgent

Botnet

  • Tengu
  • ToxNetV2

Loaders

  • KlisatuLoader
  • LpxiterDownloader
  • AzaleaDropper
  • SynkLoader

RATs

  • EdgeRAT
  • HeroinRAT
  • AccestoolesRAT
  • B4B3RAT
  • TeleRAT
  • BoratRAT
  • DriversRAT
  • GrobRAT
  • PotatoRAT
  • ArvionRAT
  • JumpscareRAT
  • XenRAT

Backdoors

  • SLEEPWALKER

Ransomware

  • SuperboyKing
  • Carso
  • OpnKey
  • Deadlock
  • RansomScare

Other

  • Detecting sandbox checker scripts
  • CapyAgent
  • MultiGoDDoS
  • Cruciferra
  • NeonVanguard
  • AWManager PUA
  • WorldProxyTool PUA
  • AzaleaC2 hacktool
  • SnakeBite
  • Pastejacking campaign
  • ToDesk RMM
  • Themida packer
  • PrefireWiper
  • AckiMiner

Written by

Izabela Komorowska

Izabela Komorowska

Technical writer

Technical writer with over nine years of experience in various tech companies, ranging from small startups to large SaaS enterprises. Enthusiastic about writing easy-to-understand customer documentation that clarifies a product’s features and benefits.

Izabela on LinkedIn