Security researchers at Fortinet FortiGuard Labs have detailed a multi-stage phishing campaign targeting individuals in Russia. The attack uses business-themed lures to trick victims into executing malicious scripts. The campaign is characterized by its sophisticated delivery mechanism, which leverages public cloud services like GitHub and Dropbox to host different stages of the attack. A key component of the attack is the use of a defense evasion tool named 'defendnot' to disable Microsoft Defender. The final payloads include the Amnesia remote access trojan (RAT) and an unspecified strain of ransomware, enabling both espionage and financial extortion.
This campaign demonstrates a 'living-off-the-trusted-land' approach, where attackers abuse legitimate services to increase the likelihood of success and complicate detection and takedown efforts. By hosting payloads on reputable services like GitHub and Dropbox, the malicious traffic is more likely to bypass network security controls that might block connections to unknown or suspicious domains. The attack does not rely on zero-day exploits, but rather on deceiving the user and manipulating built-in system tools, making user awareness and endpoint hardening critical for defense.
The attack proceeds through a carefully orchestrated chain of events:
T1566.001 - Phishing: Spearphishing AttachmentT1562.001 - Impair Defenses: Disable or Modify ToolsT1059.001 - Command and Scripting Interpreter: PowerShellThe dual payload of a RAT and ransomware creates a multi-faceted threat:
| Type | Value | Description |
|---|---|---|
| url_pattern | raw.githubusercontent.com |
Monitor for PowerShell or other scripting engines making connections to download content from GitHub. |
| url_pattern | dropbox.com/s/ |
Monitor for direct downloads of executables (.exe, .dll) from Dropbox shared links via command line tools. |
| process_name | msseces.exe |
The 'defendnot' tool reportedly manipulates this process (Microsoft Security Client) to disable Defender. Anomalous behavior from this process is suspicious. |
| registry_key | HKLM\SOFTWARE\Microsoft\Security Center\Provider\Av |
Monitor for unauthorized changes to this registry key, which tracks the status of antivirus products. |
cscript.exe/wscript.exe processes making outbound connections to GitHub and Dropbox. EDR solutions should be configured to alert on any attempts to tamper with Microsoft Defender services or registry keys. Microsoft's Tamper Protection feature, if enabled, is designed to block the actions of tools like 'defendnot'. Use D3FEND technique D3-SCA: System Call Analysis to detect unusual API calls related to the Windows Security Center.Enable security features like Microsoft Defender's Tamper Protection to prevent unauthorized disabling of security tools.
Use application control to prevent the execution of untrusted scripts downloaded from the internet.
Train users to be suspicious of unsolicited business documents and to report phishing attempts.
Block or inspect downloads of executables from non-corporate cloud storage services.
The most direct countermeasure to the 'defendnot' tool is to enable Microsoft Defender's built-in Tamper Protection feature via Group Policy or Intune. This specific hardening configuration is designed to prevent unauthorized changes to security settings, including attempts to disable real-time protection or modify security intelligence updates. By enabling this, the attack chain is broken at the defense evasion stage, as the 'defendnot' tool's API calls to disable Defender would be blocked. This ensures the endpoint security solution remains active to detect and block the subsequent download and execution of the Amnesia RAT and ransomware payloads from Dropbox.
Implement application allowlisting using a tool like Windows Defender Application Control (WDAC). This would prevent the execution of the unsigned and untrusted scripts and binaries downloaded from GitHub and Dropbox. In a properly configured allowlisting environment, even if a user is tricked into running the initial file, the subsequent PowerShell script fetched from GitHub would be blocked from executing because its hash or signer is not on the allowlist. This control moves the security posture from a reactive 'detect badness' model to a proactive 'only allow goodness' model, which is highly effective against multi-stage attacks that drop new executables onto a system.

Cybersecurity professional with over 10 years of specialized experience in security operations, threat intelligence, incident response, and security automation. Expertise spans SOAR/XSOAR orchestration, threat intelligence platforms, SIEM/UEBA analytics, and building cyber fusion centers. Background includes technical enablement, solution architecture for enterprise and government clients, and implementing security automation workflows across IR, TIP, and SOC use cases.
Help others stay informed about cybersecurity threats