AA21-076A: TrickBot Malware

This post was originally published on this site

Original release date: March 17, 2021


This Advisory uses the MITRE Adversarial Tactics, Techniques, and Common Knowledge (ATT&CK®) framework. See the ATT&CK for Enterprise for all referenced threat actor tactics and techniques.

The Cybersecurity and Infrastructure Security Agency (CISA) and Federal Bureau of Investigation (FBI) have observed continued targeting through spearphishing campaigns using TrickBot malware in North America. A sophisticated group of cybercrime actors is luring victims, via phishing emails, with a traffic infringement phishing scheme to download TrickBot.

TrickBot—first identified in 2016—is a Trojan developed and operated by a sophisticated group of cybercrime actors. Originally designed as a banking Trojan to steal financial data, TrickBot has evolved into highly modular, multi-stage malware that provides its operators a full suite of tools to conduct a myriad of illegal cyber activities.

To secure against TrickBot, CISA and FBI recommend implementing the mitigation measures described in this Joint Cybersecurity Advisory, which include blocking suspicious Internet Protocol addresses, using antivirus software, and providing social engineering and phishing training to employees.

Click here for a PDF version of this report.

Technical Details

TrickBot is an advanced Trojan that malicious actors spread primarily by spearphishing campaigns using tailored emails that contain malicious attachments or links, which—if enabled—execute malware (Phishing: Spearphishing Attachment [T1566.001], Phishing: Spearphishing Link [T1566.002]). CISA and FBI are aware of recent attacks that use phishing emails, claiming to contain proof of a traffic violation, to steal sensitive information. The phishing emails contain links that redirect to a website hosted on a compromised server that prompts the victim to click on photo proof of their traffic violation. In clicking the photo, the victim unknowingly downloads a malicious JavaScript file that, when opened, automatically communicates with the malicious actor’s command and control (C2) server to download TrickBot to the victim’s system.

Attackers can use TrickBot to:

  • Drop other malware, such as Ryuk and Conti ransomware, or
  • Serve as an Emotet downloader.[1]

TrickBot uses person-in-the-browser attacks to steal information, such as login credentials (Man in the Browser [T1185]). Additionally, some of TrickBot’s modules spread the malware laterally across a network by abusing the Server Message Block (SMB) Protocol. TrickBot operators have a toolset capable of spanning the entirety of the MITRE ATT&CK framework, from actively or passively gathering information that can be used to support targeting (Reconnaissance [TA0043]), to trying to manipulate, interrupt, or destroy systems and data (Impact [TA0040]).

TrickBot is capable of data exfiltration, cryptomining, and host enumeration (e.g., reconnaissance of Unified Extensible Firmware Interface or Basic Input/Output System [UEFI/BIOS] firmware).[2] For host enumeration, operators deliver TrickBot in modules containing a configuration file with specific tasks.

Figure 1 lays out TrickBot’s use of enterprise techniques.

Figure 1: MITRE ATT&CK enterprise techniques used by TrickBot


MITRE ATT&CK Techniques

According to MITRE, TrickBot [S0266] uses the ATT&CK techniques listed in table 1.

Table 1: TrickBot ATT&CK techniques for enterprise

Initial Access [TA0001]

Technique Title

ID Use
Phishing: Spearphishing Attachment T1566.001 TrickBot has used an email with an Excel sheet containing a malicious macro to deploy the malware.
Phishing: Spearphishing Link T1566.002

TrickBot has been delivered via malicious links in phishing emails.

Execution [TA0002]

Technique Title ID Use
Scheduled Task/Job: Scheduled Task T1053.005 TrickBot creates a scheduled task on the system that provides persistence.
Command and Scripting Interpreter: Windows Command Shell T1059.003 TrickBot has used macros in Excel documents to download and deploy the malware on the user’s machine.
Native API T1106 TrickBot uses the Windows Application Programming Interface (API) call, CreateProcessW(), to manage execution flow.
User Execution: Malicious Link T1204.001 TrickBot has sent spearphishing emails in an attempt to lure users to click on a malicious link.
User Execution: Malicious File T1204.002 TrickBot has attempted to get users to launch malicious documents to deliver its payload.

Persistence [TA0003]

Technique Title ID Use
Scheduled Task/Job: Scheduled Task T1053.005 TrickBot creates a scheduled task on the system that provides persistence.
Create or Modify System Process: Windows Service T1543.003 TrickBot establishes persistence by creating an autostart service that allows it to run whenever the machine boots.

Privilege Escalation [TA0004]

Technique Title ID Use
Scheduled Task/Job: Scheduled Task T1053.005 TrickBot creates a scheduled task on the system that provides persistence.
Process Injection: Process Hollowing T1055.012 TrickBot injects into the svchost.exe process.
Create or Modify System Process: Windows Service T1543.003 TrickBot establishes persistence by creating an autostart service that allows it to run whenever the machine boots.

 Defense Evasion [TA0005]

Technique Title ID Use
Obfuscated Files or Information T1027 TrickBot uses non-descriptive names to hide functionality and uses an AES CBC (256 bits) encryption algorithm for its loader and configuration files.
Obfuscated Files or Information: Software Packing T1027.002 TrickBot leverages a custom packer to obfuscate its functionality.
Masquerading T1036 The TrickBot downloader has used an icon to appear as a Microsoft Word document.
Process Injection: Process Hollowing T1055.012 TrickBot injects into the svchost.exe process.
Modify Registry T1112 TrickBot can modify registry entries.
Deobfuscate/Decode Files or Information T1140 TrickBot decodes the configuration data and modules.
Subvert Trust Controls: Code Signing T1553.002 TrickBot has come with a signed downloader component.
Impair Defenses: Disable or Modify Tools T1562.001 TrickBot can disable Windows Defender.

Credential Access [TA0006]

Technique Title ID Use
Input Capture: Credential API Hooking T1056.004 TrickBot has the ability to capture Remote Desktop Protocol credentials by capturing the CredEnumerateA API.
Unsecured Credentials: Credentials in Files T1552.001 TrickBot can obtain passwords stored in files from several applications such as Outlook, Filezilla, OpenSSH, OpenVPN and WinSCP. Additionally, it searches for the .vnc.lnk affix to steal VNC credentials.
Unsecured Credentials: Credentials in Registry T1552.002 TrickBot has retrieved PuTTY credentials by querying the SoftwareSimonTathamPuttySessions registry key.
Credentials from Password Stores T1555 TrickBot can steal passwords from the KeePass open-source password manager.
Credentials from Password Stores: Credentials from Web Browsers T1555.003 TrickBot can obtain passwords stored in files from web browsers such as Chrome, Firefox, Internet Explorer, and Microsoft Edge, sometimes using esentutl.

Discovery [TA0007]

Technique Tactic ID Use
System Service Discovery T1007 TrickBot collects a list of install programs and services on the system’s machine.
System Network Configuration Discovery T1016 TrickBot obtains the IP address, location, and other relevant network information from the victim’s machine.
Remote System Discovery T1018 TrickBot can enumerate computers and network devices.
System Owner/User Discovery T1033 TrickBot can identify the user and groups the user belongs to on a compromised host.
Permission Groups Discovery T1069 TrickBot can identify the groups the user on a compromised host belongs to.
System Information Discovery T1082 TrickBot gathers the OS version, machine name, CPU type, amount of RAM available from the victim’s machine.
File and Directory Discovery T1083 TrickBot searches the system for all of the following file extensions: .avi, .mov, .mkv, .mpeg, .mpeg4, .mp4, .mp3, .wav, .ogg, .jpeg, .jpg, .png, .bmp, .gif, .tiff, .ico, .xlsx, and .zip. It can also obtain browsing history, cookies, and plug-in information.
Account Discovery: Local Account T1087.001 TrickBot collects the users of the system.
Account Discovery: Email Account T1087.003 TrickBot collects email addresses from Outlook.
Domain Trust Discovery T1482 TrickBot can gather information about domain trusts by utilizing Nltest.

Collection [TA0009]

Technique Tactic ID Use
Data from Local System T1005 TrickBot collects local files and information from the victim’s local machine.
Input Capture:Credential API Hooking T1056.004 TrickBot has the ability to capture Remote Desktop Protocol credentials by capturing the CredEnumerateA API.
Person in the Browser T1185 TrickBot uses web injects and browser redirection to trick the user into providing their login credentials on a fake or modified webpage.

Command and Control [TA0011]

Technique Tactic ID Use
Fallback Channels T1008 TrickBot can use secondary command and control (C2) servers for communication after establishing connectivity and relaying victim information to primary C2 servers.
Application Layer Protocol: Web Protocols T1071.001 TrickBot uses HTTPS to communicate with its C2 servers, to get malware updates, modules that perform most of the malware logic and various configuration files.
Ingress Tool Transfer T1105 TrickBot downloads several additional files and saves them to the victim’s machine.
Data Encoding: Standard Encoding T1132.001 TrickBot can Base64-encode C2 commands.
Non-Standard Port T1571 Some TrickBot samples have used HTTP over ports 447 and 8082 for C2.
Encrypted Channel: Symmetric Cryptography T1573.001 TrickBot uses a custom crypter leveraging Microsoft’s CryptoAPI to encrypt C2 traffic.

Exfiltration [TA0010]

Technique Tactic ID Use
Exfiltration Over C2 Channel T1041 TrickBot can send information about the compromised host to a hardcoded C2 server.



CISA developed the following snort signature for use in detecting network activity associated with TrickBot activity.


alert tcp any [443,447] -> any any (msg:”TRICKBOT:SSL/TLS Server X.509 Cert Field contains ‘example.com’ (Hex)”; sid:1; rev:1; flow:established,from_server; ssl_state:server_hello; content:”|0b|example.com”; fast_pattern:only; content:”Global Security”; content:”IT Department”; pcre:”/(?:x09x00xc0xb9x3bx93x72xa3xf6xd2|x00xe2x08xffxfbx7bx53x76x3d)/”; classtype:bad-unknown; metadata:service ssl,service and-ports;)


alert tcp any any -> any $HTTP_PORTS (msg:”TRICKBOT_ANCHOR:HTTP URI GET contains ‘/anchor'”; sid:1; rev:1; flow:established,to_server; content:”/anchor”; http_uri; fast_pattern:only; content:”GET”; nocase; http_method; pcre:”/^/anchor_?.{3}/[w_-]+.[A-F0-9]+/?$/U”; classtype:bad-unknown; priority:1; metadata:service http;)


alert tcp any $SSL_PORTS -> any any (msg:”TRICKBOT:SSL/TLS Server X.509 Cert Field contains ‘C=XX, L=Default City, O=Default Company Ltd'”; sid:1; rev:1; flow:established,from_server; ssl_state:server_hello; content:”|31 0b 30 09 06 03 55 04 06 13 02|XX”; nocase; content:”|31 15 30 13 06 03 55 04 07 13 0c|Default City”; nocase; content:”|31 1c 30 1a 06 03 55 04 0a 13 13|Default Company Ltd”; nocase; content:!”|31 0c 30 0a 06 03 55 04 03|”; classtype:bad-unknown; reference:url,www.virustotal.com/gui/file/e9600404ecc42cf86d38deedef94068db39b7a0fd06b3b8fb2d8a3c7002b650e/detection; metadata:service ssl;)


alert tcp any any -> any $HTTP_PORTS (msg:”TRICKBOT:HTTP Client Header contains ‘boundary=Arasfjasu7′”; sid:1; rev:1; flow:established,to_server; content:”boundary=Arasfjasu7|0d 0a|”; http_header; content:”name=|22|proclist|22|”; http_header; content:!”Referer”; content:!”Accept”; content:”POST”; http_method; classtype:bad-unknown; metadata:service http;)


alert tcp any any -> any $HTTP_PORTS (msg:”TRICKBOT:HTTP Client Header contains ‘User-Agent|3a 20|WinHTTP loader/1.'”; sid:1; rev:1; flow:established,to_server; content:”User-Agent|3a 20|WinHTTP loader/1.”; http_header; fast_pattern:only; content:”.png|20|HTTP/1.”; pcre:”/^Hostx3ax20(?:d{1,3}.){3}d{1,3}(?:x3ad{2,5})?$/mH”; content:!”Accept”; http_header; content:!”Referer|3a 20|”; http_header; classtype:bad-unknown; metadata:service http;)


alert tcp any $HTTP_PORTS -> any any (msg:”TRICKBOT:HTTP Server Header contains ‘Server|3a 20|Cowboy'”; sid:1; rev:1; flow:established,from_server; content:”200″; http_stat_code; content:”Server|3a 20|Cowboy|0d 0a|”; http_header; fast_pattern; content:”content-length|3a 20|3|0d 0a|”; http_header; file_data; content:”/1/”; depth:3; isdataat:!1,relative; classtype:bad-unknown; metadata:service http;)


alert tcp any any -> any $HTTP_PORTS (msg:”TRICKBOT:HTTP URI POST contains C2 Exfil”; sid:1; rev:1; flow:established,to_server; content:”Content-Type|3a 20|multipart/form-data|3b 20|boundary=——Boundary”; http_header; fast_pattern; content:”User-Agent|3a 20|”; http_header; distance:0; content:”Content-Length|3a 20|”; http_header; distance:0; content:”POST”; http_method; pcre:”/^/[a-z]{3}d{3}/.+?.[A-F0-9]{32}/d{1,3}//U”; pcre:”/^Hostx3ax20(?:d{1,3}.){3}d{1,3}$/mH”; content:!”Referer|3a|”; http_header; classtype:bad-unknown; metadata:service http;)


alert tcp any any -> any $HTTP_PORTS (msg:”HTTP URI GET/POST contains ‘/56evcxv’ (Trickbot)”; sid:1; rev:1; flow:established,to_server; content:”/56evcxv”; http_uri; fast_pattern:only; classtype:bad-unknown; metadata:service http;)


alert icmp any any -> any any (msg:”TRICKBOT_ICMP_ANCHOR:ICMP traffic conatins ‘hanc'”; sid:1; rev:1; itype:8; content:”hanc”; offset:4; fast_pattern; classtype:bad-unknown;)


alert tcp any any -> any $HTTP_PORTS (msg:”HTTP Client Header contains POST with ‘host|3a 20|*.onion.link’ and ‘data=’ (Trickbot/Princess Ransomeware)”; sid:1; rev:1; flow:established,to_server; content:”POST”; nocase; http_method; content:”host|3a 20|”; http_header; content:”.onion.link”; nocase; http_header; distance:0; within:47; fast_pattern; file_data; content:”data=”; distance:0; within:5; classtype:bad-unknown; metadata:service http;)


alert tcp any any -> any $HTTP_PORTS (msg:”HTTP Client Header contains ‘host|3a 20|tpsci.com’ (trickbot)”; sid:1; rev:1; flow:established,to_server; content:”host|3a 20|tpsci.com”; http_header; fast_pattern:only; classtype:bad-unknown; metadata:service http;)


CISA and FBI recommend that network defenders—in federal, state, local, tribal, territorial governments, and the private sector—consider applying the following best practices to strengthen the security posture of their organization’s systems. System owners and administrators should review any configuration changes prior to implementation to avoid negative impacts.

  • Provide social engineering and phishing training to employees.
  • Consider drafting or updating a policy addressing suspicious emails  that specifies users must report all suspicious emails to the security and/or IT departments.
  • Mark external emails with a banner denoting the email is from an external source to assist users in detecting spoofed emails.
  • Implement Group Policy Object and firewall rules.
  • Implement an antivirus program and a formalized patch management process.
  • Implement filters at the email gateway and block suspicious IP addresses at the firewall.
  • Adhere to the principle of least privilege.
  • Implement a Domain-Based Message Authentication, Reporting & Conformance validation system.
  • Segment and segregate networks and functions.
  • Limit unnecessary lateral communications between network hoses, segments and devices.
  • Consider using application allowlisting technology on all assets to ensure that only authorized software executes, and all unauthorized software is blocked from executing on assets. Ensure that such technology only allows authorized, digitally signed scripts to run on a system.
  • Enforce multi-factor authentication.
  • Enable a firewall on agency workstations configured to deny unsolicited connection requests.
  • Disable unnecessary services on agency workstations and servers.
  • Implement an Intrusion Detection System, if not already used, to detect C2 activity and other potentially malicious network activity
  • Monitor web traffic. Restrict user access to suspicious or risky sites.
  • Maintain situational awareness of the latest threats and implement appropriate access control lists.
  • Disable the use of SMBv1 across the network and require at least SMBv2 to harden systems against network propagation modules used by TrickBot.
  • Visit the MITRE ATT&CK Techniques pages (linked in table 1 above) for additional mitigation and detection strategies.
  • See CISA’s Alert on Technical Approaches to Uncovering and Remediating Malicious Activity for more information on addressing potential incidents and applying best practice incident response procedures.

For additional information on malware incident prevention and handling, see the National Institute of Standards and Technology Special Publication 800-83, Guide to Malware Incident Prevention and Handling for Desktops and Laptops.




  • March 17, 2021: Initial Version

This product is provided subject to this Notification and this Privacy & Use policy.

SecretStore Release Candidate 3

This post was originally published on this site

The SecretStore release candidate 3 (RC3) module is now available on the PowerShell Gallery. This contains an exciting new feature which allows users to non-interactively create, and configure a SecretStore. This feature was added to support CI systems and other automated scenarios.

SecretStore is an extension vault module, for PowerShell SecretManagement, which works over all supported PowerShell platforms on Windows, Linux, and macOS. For more context on this module and the SecretManagement module refer to the previous blog posts:

For more context on these modules check out these previous blog posts:

Before installing this module, please uninstall the current preview versions of the module and restart your PowerShell session.

To install these updates run the following commands:

Uninstall-Module Microsoft.PowerShell.SecretStore -Force
# Restart your PowerShell session
Install-Module -Name Microsoft.PowerShell.SecretStore -Repository PSGallery
Register-SecretVault -Name SecretStore -ModuleName Microsoft.PowerShell.SecretStore -DefaultVault -AllowClobber

SecretStore Updates

Previously, Set-SecretStoreConfiguration required manual password confirmation to make changes to SecretStore configuration. This update adds a -Password parameter to Set-SecretStoreConfiguration to endable automated creation and configuration of a SecretVault.

Breaking Change

  • The -Force parameter was removed from the Set-SecretStoreConfiguration command, and instead the -Confirm:$false should be used to suppress PowerShell confirmation prompting in automation scripts.

New Feature

  • Set-SecretStoreConfiguration command now takes a -Password parameter so that there is no need to prompt for a password

How to non-interactively create and configure a SecretStore

This is an example of automation script that installs and configures the Microsoft.PowerShell.SecretStore module without user prompting. The configuration requires a password and sets user interaction to None, so that SecretStore will never prompt the user. The configuration also requires a password, and the password is passed in as a SecureString object. The -Confirm:false parameter is used so that PowerShell will not prompt for confirmation.

The SecretStore password must be provided in a secure fashion. Here the password is being imported from an encrypted file using Windows Data Protection API, but this is a Windows only solution. Another option is to use a CI system mechanism such as secure variables.

Next, the SecretManagement module is installed and the SecretStore module registered so that the SecretStore secrets can be managed.

The Unlock-SecretStore cmdlet is used to unlock the SecretStore for this session. The password timeout was configured for 1 hour and SecretStore will remain unlocked in the session for that amount of time, after which it will need to be unlocked again before secrets can be accessed.

Install-Module -Name Microsoft.PowerShell.SecretStore -Repository PSGallery -Force
$password = Import-CliXml -Path $securePasswordPath

Set-SecretStoreConfiguration -Scope CurrentUser -Authentication Password -PasswordTimeout 3600 -Interaction None -Password $password -Confirm:$false

Install-Module -Name Microsoft.PowerShell.SecretManagement -Repository PSGallery -Force
Register-SecretVault -Name SecretStore -ModuleName Microsoft.PowerShell.SecretStore -DefaultVault

Unlock-SecretStore -Password $password

General Availability (GA)

This is a “go live” release, which means that we feel that this RC is feature complete and supported in production.

If no bugs are identified through this release, we will increment the versioning and declare the module as GA in late March. If any high-risk bugs are identified we will continue to release RCs until the quality bar is met for a GA release.

Feedback and Support

Community feedback has been essential to the iterative development of these modules. Thank you to everyone who has contributed issues, and feedback thus far! To file issues or get support for the SecretManagement interface or vault development experience please use the SecretManagement repository. For issues which pertain specifically to the SecretStore and its cmdlet interface please use the SecretStore repository.

Sydney Smith

PowerShell Team



The post SecretStore Release Candidate 3 appeared first on PowerShell Team.

Updating help for the PSReadLine module in Windows PowerShell 5.1

This post was originally published on this site

Back in November of 2020 I
posted instructions
about how to work around the problem of updating the help for the PSReadLine module.

The version of the PSReadline module that shipped in Windows PowerShell 5.1 used a lowercase
letter in the name. The name of the module was changed for the release of PowerShell 6. It now uses
a capital L in the name. And even after following the instructions in the previous post, you still
get the same error when you try to update help for Windows PowerShell 5.1.

Failed to update Help for the module(s) ‘PSReadline

The root problem in Windows PowerShell 5.1 is that the module is automatically loaded at startup
using the PSReadline name (with the lowercase l). Use the following steps workaround this
problem and update the help.

  1. Close all powershell.exe processes. Then rename the PSReadline module folder name to
  2. Open an powershell.exe session in Administrator privilege.
  3. Run Remove-Module PSReadLine to remove the auto-loaded PSReadLine module.
  4. Run Import-Module PSReadLine (use PSReadLine with capital L).
  5. Then run Update-Help -Module PSReadLine -Force

These steps load the module in your session using the new name, which allows Update-Help to use the
correct name to find the help content.

The post Updating help for the PSReadLine module in Windows PowerShell 5.1 appeared first on PowerShell Team.

Announcing PSReadLine 2.2-beta.2 with Dynamic help

This post was originally published on this site

We are pleased to announce the new feature Dynamic Help in
PSReadLine 2.2.0-beta.2.
Features like TabCompletetion and the Help system make PowerShell easier to use. Expanding on that
theme, Dynamic Help provides just-in-time help at the command line. This experience allows you to
stay focused on your work in the terminal without losing your place typing on the pipeline.

Supported Releases

Dynamic Help is implemented and presented through the PSReadLine module. In an effort to bring the
benefits of Dynamic Help to the most users, Dynamic Help is supported down-level to Windows
PowerShell 5.1.

PSReadLine 2.2.0-beta2, including Dynamic Help, is supported in the following versions:

  • Windows PowerShell 5.1
  • PowerShell 7.0+

Install PSReadLine with Dynamic Help

To receive the benefits of Dynamic Help, download and install the PSReadLine module from PSGallery.

Install-Module PSReadLine -AllowPrerelease

Getting Cmdlet Help

Dynamic Help provides a view of full cmdlet help shown on an alternative screen buffer using a cross
platform Pager utility. PSReadLine maps the function ShowCommandHelp to the F1 key.

  • When the cursor is at the end of a fully expanded cmdlet, pressing F1 displays the help for
    that cmdlet.
  • When the cursor is at the end of a fully expanded parameter, pressing F1 displays the help
    beginning at the parameter.

Image DynHelp1

The Pager in PSReadLine supports viewing content one screen at a time. To navigate the displayed
help content, press the up and down arrow keys to scroll the screen. Pressing Q exits the
alternative screen buffer and returns to the current cursor location in the the primary screen

The Pager is a work in progress and we welcome your feedback and suggestions. To provide feedback
and for more information, see the Pager repository on GitHub.

Focused Parameter Help

Pressing the Alt-h key combination provides dynamic help for parameters. The help is shown below
the current command line like MenuComplete. The cursor must be at the end of the fully-expanded
parameter name when you press the Alt-h key.

Image DynHelp Alt H


The PSReadLine function ShowParameterHelp is bound to Alt-h.
The Alt bindings do not currently work on macOS as detailed in this
issue. Mac users will need to map
ShowParameterHelp function to an available key combination.

Set-PSReadLineKeyHandler -chord "Ctrl-l" -Function ShowParameterHelp

Selection of Arguments

To rapidly select and change the arguments of a cmdlet without disturbing your syntax, press
Alt-a. Selection of arguments is scoped within a script block. Based on the cursor position, it
searches from the innermost script block to the outmost script block, and stops when it finds any
arguments in a script block scope.

Image dynhelpA


The PSReadLine function SelectCommandArgument is bound to
Alt-a. The Alt bindings do not currently work on macOS as detailed in this
issue. Mac users will need to map
SelectCommandArgument function to an available key combination.

Set-PSReadLineKeyHandler -chord "Ctrl-k" -function SelectCommandArgument

Feedback and Support

Community feedback is essential to the iterative development of PSReadLine and features like Dynamic
Help. Thank you for your continued support and engagement. To file issues for PSReadLine or Dynamic
Help, please visit the PSReadLine repository.

Jason Helmick

PowerShell Team

The post Announcing PSReadLine 2.2-beta.2 with Dynamic help appeared first on PowerShell Team.

AA21-062A: Mitigate Microsoft Exchange Server Vulnerabilities

This post was originally published on this site

Original release date: March 3, 2021


Cybersecurity and Infrastructure Security (CISA) partners have observed active exploitation of vulnerabilities in Microsoft Exchange Server products. Successful exploitation of these vulnerabilities allows an unauthenticated attacker to execute arbitrary code on vulnerable Exchange Servers, enabling the attacker to gain persistent system access, as well as access to files and mailboxes on the server and to credentials stored on that system. Successful exploitation may additionally enable the attacker to compromise trust and identity in a vulnerable network. Microsoft released out-of-band patches to address vulnerabilities in Microsoft Exchange Server. The vulnerabilities impact on-premises Microsoft Exchange Servers and are not known to impact Exchange Online or Microsoft 365 (formerly O365) cloud email services.

This Alert includes both tactics, techniques and procedures (TTPs) and the indicators of compromise (IOCs) associated with this malicious activity. To secure against this threat, CISA recommends organizations examine their systems for the TTPs and use the IOCs to detect any malicious activity. If an organization discovers exploitation activity, they should assume network identity compromise and follow incident response procedures. If an organization finds no activity, they should apply available patches immediately and implement the mitigations in this Alert.

Click here for IOCs in STIX format.

Technical Details

Microsoft has released out-of-band security updates to address four vulnerabilities in Exchange Server:

  • CVE-2021-26855 allows an unauthenticated attacker to send arbitrary HTTP requests and authenticate as the Exchange Server. The vulnerability exploits the Exchange Control Panel (ECP) via a Server-Side Request Forgery (SSRF). This would also allow the attacker to gain access to mailboxes and read sensitive information.
  • CVE-2021-26857, CVE-2021-26858, and CVE-2021-27065 allow for remote code execution.  
    • CVE-2021-26858 and CVE-2021-27065 are similar post-authentication arbitrary write file vulnerabilities in Exchange. An attacker, authenticated either by using CVE-2021-26855 or via stolen admin credentials, could write a file to any path on the server.

    • CVE-2021-26857 is an insecure deserialization vulnerability in the Unified Messaging service. An attacker, authenticated either by using CVE-2021-26855 or via stolen admin credentials, could execute arbitrary code as SYSTEM on the Exchange Server.

  • To locate a possible compromise of these CVEs, we encourage you to read the Microsoft Advisory.

It is possible for an attacker, once authenticated to the Exchange server, to gain access to the Active Directory environment and download the Active Directory Database.

Tactics, Techniques and Procedures

The majority of the TTPs in this section are sourced from a blog post from Veloxity, a third party cybersecurity firm. Note: the United States Government does not endorse any commercial product or service, including any subjects of analysis. Any reference to specific commercial products, processes, or services by service mark, trademark, manufacturer, or otherwise, does not constitute or imply their endorsement, recommendation, or favoring by the United States Government.

Veloxity has observed the following files as targets of HTTP POST requests:

  • /owa/auth/Current/themes/resources/logon.css
  • /owa/auth/Current/themes/resources/owafont_ja.css
  • /owa/auth/Current/themes/resources/lgnbotl.gif
  • /owa/auth/Current/themes/resources/owafont_ko.css
  • /owa/auth/Current/themes/resources/SegoeUI-SemiBold.eot
  • /owa/auth/Current/themes/resources/SegoeUI-SemiLight.ttf
  • /owa/auth/Current/themes/resources/lgnbotl.gif

Administrators should search the ECP server logs for the following string (or something similar):


The logs can be found at <exchange install path>LoggingECPServer.

To determine possible webshell activity, administrators should search for aspx files in the following paths:

  • inetpubwwwrootaspnet_client (any .aspx file under this folder or sub folders)
  • <exchange install path>FrontEndHttpProxyecpauth (any file besides TimeoutLogoff.aspx)
  • <exchange install path>FrontEndHttpProxyowaauth (any file or modified file that is not part of a standard install)
  • <exchange install path>FrontEndHttpProxyowaauthCurrent (any aspx file in this folder or subfolders)
  • <exchange install path>FrontEndHttpProxyowaauth<folder with version number> (any aspx file in this folder or subfolders)

Administrators should search in the /owa/auth/Current directory for the following non-standard web log user-agents. These agents may be useful for incident responders to look at to determine if further investigation is necessary.

These should not be taken as definitive IOCs:

  • DuckDuckBot/1.0;+(+http://duckduckgo.com/duckduckbot.html)
  • facebookexternalhit/1.1+(+http://www.facebook.com/externalhit_uatext.php)
  • Mozilla/5.0+(compatible;+Baiduspider/2.0;++http://www.baidu.com/search/spider.html)
  • Mozilla/5.0+(compatible;+Bingbot/2.0;++http://www.bing.com/bingbot.htm)
  • Mozilla/5.0+(compatible;+Googlebot/2.1;++http://www.google.com/bot.html
  • Mozilla/5.0+(compatible;+Konqueror/3.5;+Linux)+KHTML/3.5.5+(like+Gecko)+(Exabot-Thumbnails)
  • Mozilla/5.0+(compatible;+Yahoo!+Slurp;+http://help.yahoo.com/help/us/ysearch/slurp)
  • Mozilla/5.0+(compatible;+YandexBot/3.0;++http://yandex.com/bots)
  • Mozilla/5.0+(X11;+Linux+x86_64)+AppleWebKit/537.36+(KHTML,+like+Gecko)+Chrome/51.0.2704.103+Safari/537.36

Veloxity observed these user-agents in conjunction with exploitation to /ecp/ URLs:

  • ExchangeServicesClient/
  • python-requests/2.19.1
  • python-requests/2.25.1

These user-agents were also observed having connections to post-exploitation web-shell access:

  • antSword/v2.1
  • Googlebot/2.1+(+http://www.googlebot.com/bot.html)
  • Mozilla/5.0+(compatible;+Baiduspider/2.0;++http://www.baidu.com/search/spider.html)

As with the non-standard user-agents, responders can examine internet information services (IIS) logs from Exchange Servers to identify possible historical activity. Also, as with the non-standard user agents, these should not be taken as definitive IOCs:

  • POST /owa/auth/Current/
  • POST /ecp/default.flt
  • POST /ecp/main.css
  • POST /ecp/<single char>.js

Volexity has seen attackers leverage the following IP addresses. Although these are tied to virtual private servers (VPSs) servers and virtual private networks (VPNs), responders should investigate these IP addresses on their networks and act accordingly:


Volexity has also provided the following YARA signatures that can be run within your network to assist in finding signs of a compromise.

rule webshell_aspx_simpleseesharp : Webshell Unclassified
        author = “threatintel@volexity.com”
        date = “2021-03-01”
        description = “A simple ASPX Webshell that allows an attacker to write further files to disk.”
        hash = “893cd3583b49cb706b3e55ecb2ed0757b977a21f5c72e041392d1256f31166e2”
        $header = “<%@ Page Language=”C#” %>”
        $body = “<% HttpPostedFile thisFile = Request.Files[0];thisFile.SaveAs(Path.Combine”
        $header at 0 and
        $body and
        filesize < 1KB
rule webshell_aspx_reGeorgTunnel : Webshell Commodity
        author = “threatintel@volexity.com”
        date = “2021-03-01”
        description = “A variation on the reGeorg tunnel webshell”
        hash = “406b680edc9a1bb0e2c7c451c56904857848b5f15570401450b73b232ff38928”
        reference = “https://github.com/sensepost/reGeorg/blob/master/tunnel.aspx”
        $s1 = “System.Net.Sockets”
        $s2 = “System.Text.Encoding.Default.GetString(Convert.FromBase64String(StrTr(Request.Headers.Get”
        // a bit more experimental
        $t1 = “.Split(‘|’)”
        $t2 = “Request.Headers.Get”
        $t3 = “.Substring(“
        $t4 = “new Socket(“
        $t5 = “IPAddress ip;”
        all of ($s*) or
        all of ($t*)
rule webshell_aspx_sportsball : Webshell Unclassified
        author = “threatintel@volexity.com”
        date = “2021-03-01”
        description = “The SPORTSBALL webshell allows attackers to upload files or execute commands on the system.”
        hash = “2fa06333188795110bba14a482020699a96f76fb1ceb80cbfa2df9d3008b5b0a”
        $uniq1 = “HttpCookie newcook = new HttpCookie(”fqrspt”, HttpContext.Current.Request.Form”
        $uniq2 = “ZN2aDAB4rXsszEvCLrzgcvQ4oi5J1TuiRULlQbYwldE=”
        $var1 = “Result.InnerText = string.Empty;”
        $var2 = “newcook.Expires = DateTime.Now.AddDays(”
        $var3 = “System.Diagnostics.Process process = new System.Diagnostics.Process();”
        $var4 = “process.StandardInput.WriteLine(HttpContext.Current.Request.Form[””
        $var5 = “else if (!string.IsNullOrEmpty(HttpContext.Current.Request.Form[””
        $var6 = “<input type=”submit” value=”Upload” />”
        any of ($uniq*) or
        all of ($var*)

A list of web shell hashes have also been provided by Microsoft:

  • b75f163ca9b9240bf4b37ad92bc7556b40a17e27c2b8ed5c8991385fe07d17d0
  • 097549cf7d0f76f0d99edf8b2d91c60977fd6a96e4b8c3c94b0b1733dc026d3e
  • 2b6f1ebb2208e93ade4a6424555d6a8341fd6d9f60c25e44afe11008f5c1aad1
  • 65149e036fff06026d80ac9ad4d156332822dc93142cf1a122b1841ec8de34b5
  • 511df0e2df9bfa5521b588cc4bb5f8c5a321801b803394ebc493db1ef3c78fa1
  • 4edc7770464a14f54d17f36dc9d0fe854f68b346b27b35a6f5839adf1f13f8ea
  • 811157f9c7003ba8d17b45eb3cf09bef2cecd2701cedb675274949296a6a183d
  • 1631a90eb5395c4e19c7dbcbf611bbe6444ff312eb7937e286e4637cb9e72944

Note: this is not an all-inclusive list of indicators of compromise and threat actors have been known to use short-term leased IP addresses that change very frequently. Organizations that do not locate any of the IOCs in this Alert within your network traffic, may nevertheless have been compromised. CISA recommendations following the guidance located in the Microsoft Advisory to check your servers for any signs of a compromise.  

Conduct Forensic Analysis

Should your organization see evidence of compromise, your incident response should begin with conducting forensic analysis to collect artifacts and perform triage. Please see the following list of recommendations on how to conduct forensic analysis using various tools.

Although the following free tools are not endorsed by the Federal Government, incident responders commonly use them to perform forensics.

While collecting artifacts to perform triage, use processes and tools that minimize the alteration of the data being collected and that minimize impact to the operating system itself.

Ideally, during data collection, store the data on removable/external media and, when possible, run the artifact collection tools from the same media.

Key artifacts for triage that should be collected:

  • Memory
  • All registry hives
  • All windows event logs
  • All web logs

Memory can be collected with a variety of open source tools (e.g., FTK Imager by AccessData, Ram Capture by Belkasoft).

Registry and Windows Event logs can be collected with a variety of open source tools as well (e.g., FTK_Imager, Kroll Artifact Parser And Extractor [KAPE]).

Web logs can also be collected with a variety of open source tools (e.g. FTK Imager).

Windows Artifact Collection Guide

Execute the following steps in order.

1) Download the latest FTK Imager from https://accessdata.com/product-download/.

  • Note: Ensure your review of and compliance with the applicable license associated with the product referenced, which can be found in the product’s User Guide. The United States Government does not endorse any commercial product or service, including any subjects of analysis. Any reference to specific commercial products, processes, or services by service mark, trademark, manufacturer, or otherwise, does not constitute or imply their endorsement, recommendation, or favoring by the United States Government.

2) Collect memory from live system using FTK Imager. See Memory Capture with FTK Imager.pdf for instructions. Note: Download and copy “FTK Imager” folder to an external drive. Run FTK Imager.exe from the FTK Imager folder from external drive. Wait until memory collect is complete before proceeding to step 2.

3) Collect important system artifacts using KAPE. See KAPE Collection Procedure. Note: Download KAPE from a separate system; do not download KAPE to the target system. Run KAPE from external drive.

4) Collect disk image using FTK Imager. See Live Image with FTK Imager.pdf for instructions. Note: Run FTK Imager.exe from the “FTK Imager” folder from external drive.

Memory Capture with FTK Imager

1) Open FTK Imager. Log into the system with Administrator privileges and launch “FTK Imager.”

2) Open “Capture Memory.” Select “Capture Memory…” from the File menu.

Figure 1: FTK Imager – Capture Memory Command

3) Select Path and Filenames. On the window that appears, use the “Browse” button to identify the destination of the memory capture. Save the memory capture to an external device and not the main hard drive of the system. Doing so will prevent the saved file from overwriting any dataspace on the system.

  • Name the destination file with a descriptive name (i.e., hostname of the system).
  • Select the box “Include pagefile” and provide a name of the pagefile that is descriptive of the system.
  • Do not select “Create AD1 file.”

Figure 2: FTK Imager – Memory Capture

4) Capture Memory. Click on “Capture Memory” to begin the capture process. The process will take several minutes depending on the size of the pagefile and the amount of memory on the system.

Figure 3: FTK Imager – Memory Capture

KAPE Collection Procedure [1]

1) Download KAPE from https://www.kroll.com/en/services/cyber-risk/investigate-and-respond/kroll-artifact-parser-extractor-kape.

2) Disable any antivirus or host protection mechanisms that prevent execution from removable media, or data loss prevention (DLP) mechanisms that restrict utilization of removable media.

  • Enable antivirus and host protection once this process is completed.

3) Unzip Kape.zip and run gkape.exe as admin from your removable media

4) Target source should be the drive on which the OS resides, typically C:.

5) Target destination should be an external drive folder, not the same drive as the Target source. If available, use an external hard drive or flash drive.

  • A KAPE execution with these parameters will typically produce output artifacts with a total size of 1-25 GB.
  • If you are going to be running KAPE on different machines and want to save to the same drive, ensure the Target destination folder is unique for each execution of KAPE.

6) Uncheck Flush checkbox (it is checked natively).

7) Check Add %d and Add %m checkboxes.

8) Select ALL checkboxes to ensure KAPE will target all available data that it is capable of targeting. This takes some time; use the down arrow and space bar to move through the list quickly.

9) Check Process VSCs checkbox.

10) Select Zip radio button and add Base name TargetOutput.

11) Ensure Deduplicate checkbox is checked (it is checked natively).

  • At the bottom you should now see a large Current command line, similar to:
.kape.exe –tsource C: –tdest E:%d%m –tflush –target !BasicCollection,!SANS_Triage,Avast,AviraAVLogs,Bitdefender,ComboFix,ESET,FSecure,HitmanPro,Malwarebytes, McAfee,McAfee_ePO,RogueKiller,SentinelOne,Sophos,SUPERAntiSpyware,Symantec_AV_Logs,TrendMicro,VIPRE, Webroot,WindowsDefender,Ammyy,AsperaConnect,BoxDrive,CiscoJabber,CloudStorage,ConfluenceLogs,Discord, Dropbox, Exchange,ExchangeClientAccess,ExchangeTransport,FileZilla,GoogleDrive,iTunesBackup,JavaWebCache,Kaseya,LogMeIn,Notepad++, OneDrive,OutlookPSTOST,ScreenConnect,Skype,TeamViewerLogs,TeraCopy,VNCLogs, Chrome,ChromeExtensions,Edge,Firefox,InternetExplorer,WebBrowsers,ApacheAccessLog,IISLogFiles,ManageEngineLogs, MSSQLErrorLog,NGINXLogs,PowerShellConsole,KapeTriage,MiniTimelineCollection,RemoteAdmin, VirtualDisks, Gigatribe,TorrentClients,Torrents,$Boot,$J,$LogFile,$MFT,$SDS,$T,Amcache,ApplicationEvents,BCD,CombinedLogs, EncapsulationLogging,EventLogs,EventLogs-RDP,EventTraceLogs, EvidenceOfExecution,FileSystem,GroupPolicy,LinuxOnWindowsProfileFiles,LnkFilesAndJumpLists,LogFiles,MemoryFiles, MOF,OfficeAutosave,OfficeDocumentCache,Prefetch,RDPCache,RDPLogs,RecentFileCache,Recycle, RecycleBin, RecycleBinContent,RecycleBinMetadata,RegistryHives,RegistryHivesSystem,RegistryHivesUser,ScheduledTasks,SDB, SignatureCatalog,SRUM,StartupInfo,Syscache,ThumbCache,USBDevicesLogs,WBEM,WER,WindowsFirewall,  WindowsIndexSearch,WindowsNotifcationsDB,WindowsTimeline,XPRestorePoints –vss –zip TargetOutput –gui
  • In the bottom right corner hit the Execute! Button.
  • Screenshot below shows gkape.exe during execution, you will also see a command window execute. Note: KAPE usually takes less than 20 minutes to complete on a workstation; if it is taking significantly longer there may be an issue.

Figure 4: gkape.exe screenshot


CISA strongly recommends organizations read Microsoft’s advisory and security blog post for more information on how to look for this malicious activity and apply critical patches as soon as possible.

If patching is not an immediate option, there are other mitigation options available. However, these options should only be used as a temporary solution, not a replacement for patching.  CISA recommends limiting or blocking external access to internet-facing Exchange Servers via the following:

  • Restrict untrusted connections to port 443, or set up a VPN to separate the Exchange Server from external access; note that this will not prevent an adversary from exploiting the vulnerability if the attacker is already in your network.
  • Block external access to on-premise Exchange:
    • Restrict external access to OWA URL: /owa/
    • Restrict external access to Exchange Admin Center (EAC) aka Exchange Control Panel (ECP): url /ecp/.

CISA would like to thank Microsoft and Volexity for their contributions to this alert.





  • March 3, 2021: Initial Version

This product is provided subject to this Notification and this Privacy & Use policy.