MITRE ATT&CK® techniques for Windows

Adversaries may utilize Direct Prompt Injection, Indirect Prompt Injection or other methods to instruct enterprise AI agents to exfiltrate data.[1][2] When enterprise systems incorporate the use of AI assistants or other AI agents, adversaries may abuse these systems to weaponize sanctioned tools available to the agent.[3]

As an AI agent processes queries or ingests data, an injected prompt can cause the agent to use its tool connections in unauthorized or unintended ways. When AI agents are induced to exfiltrate data, other exfiltration techniques likely apply as well to transfer the information out of the network, such as Exfiltration to Code Repository or Exfiltration Over Web Service.[4][5]

Adversaries may poison Address Resolution Protocol (ARP) caches to position themselves between the communication of two or more networked devices. This activity may be used to enable follow-on behaviors such as Network Sniffing or Transmitted Data Manipulation.

The ARP protocol is used to resolve IPv4 addresses to link layer addresses, such as a media access control (MAC) address.[1] Devices in a local network segment communicate with each other by using link layer addresses. If a networked device does not have the link layer address of a particular networked device, it may send out a broadcast ARP request to the local network to translate the IP address to a MAC address. The device with the associated IP address directly replies with its MAC address. The networked device that made the ARP request will then use as well as store that information in its ARP cache.

An adversary may passively wait for an ARP request to poison the ARP cache of the requesting device. The adversary may reply with their MAC address, thus deceiving the victim by making them believe that they are communicating with the intended networked device. For the adversary to poison the ARP cache, their reply must be faster than the one made by the legitimate IP address owner. Adversaries may also send a gratuitous ARP reply that maliciously announces the ownership of a particular IP address to all the devices in the local network segment.

The ARP protocol is stateless and does not require authentication. Therefore, devices may wrongly add or update the MAC address of the IP address in their ARP cache.[2][3]

Adversaries may use ARP cache poisoning as a means to intercept network traffic. This activity may be used to collect and/or relay data such as credentials, especially those sent over an insecure, unencrypted protocol.[2]

Adversaries may reveal credentials of accounts that have disabled Kerberos preauthentication by Password Cracking Kerberos messages.[1]

Preauthentication offers protection against offline Password Cracking. When enabled, a user requesting access to a resource initiates communication with the Domain Controller (DC) by sending an Authentication Server Request (AS-REQ) message with a timestamp that is encrypted with the hash of their password. If and only if the DC is able to successfully decrypt the timestamp with the hash of the user’s password, it will then send an Authentication Server Response (AS-REP) message that contains the Ticket Granting Ticket (TGT) to the user. Part of the AS-REP message is signed with the user’s password.[2]

For each account found without preauthentication, an adversary may send an AS-REQ message without the encrypted timestamp and receive an AS-REP message with TGT data which may be encrypted with an insecure algorithm such as RC4. The recovered encrypted data may be vulnerable to offline Password Cracking attacks similarly to Kerberoasting and expose plaintext credentials. [1][3]

An account registered to a domain, with or without special privileges, can be abused to list all domain accounts that have preauthentication disabled by utilizing Windows tools like PowerShell with an LDAP filter. Alternatively, the adversary may send an AS-REQ message for each user. If the DC responds without errors, the account does not require preauthentication and the AS-REP message will already contain the encrypted data. [1][3]

Cracked hashes may enable Persistence, Privilege Escalation, and Lateral Movement via access to Valid Accounts.[4]

Adversaries may circumvent mechanisms designed to control privilege elevation to gain higher-level permissions. Most modern systems contain native elevation control mechanisms that are intended to limit privileges that a user can perform on a machine. Authorization has to be granted to specific users in order to perform tasks that can be considered of higher risk.[1][2] An adversary can perform several methods to take advantage of built-in control mechanisms in order to escalate privileges on a system.[3][4]

Adversaries may modify access tokens to operate under a different user or system security context to perform actions and bypass access controls. Windows uses access tokens to determine the ownership of a running process. A user can manipulate access tokens to make a running process appear as though it is the child of a different process or belongs to someone other than the user that started the process. When this occurs, the process also takes on the security context associated with the new token.

An adversary can use built-in Windows API functions to copy access tokens from existing processes; this is known as token stealing. These token can then be applied to an existing process (i.e. Token Impersonation/Theft) or used to spawn a new process (i.e. Create Process with Token). An adversary must already be in a privileged user context (i.e. administrator) to steal a token. However, adversaries commonly use token stealing to elevate their security context from the administrator level to the SYSTEM level. An adversary can then use a token to authenticate to a remote system as the account for that token if the account has appropriate permissions on the remote system.[1]

Any standard user can use the runas command, and the Windows API functions, to create impersonation tokens; it does not require access to an administrator account. There are also other mechanisms, such as Active Directory fields, that can be used to modify access tokens.

Adversaries may establish persistence and/or elevate privileges by executing malicious content triggered by accessibility features. Windows contains accessibility features that may be launched with a key combination before a user has logged in (ex: when the user is on the Windows logon screen). An adversary can modify the way these programs are launched to get a command prompt or backdoor without logging in to the system.

Two common accessibility programs are C:\Windows\System32\sethc.exe, launched when the shift key is pressed five times and C:\Windows\System32\utilman.exe, launched when the Windows + U key combination is pressed. The sethc.exe program is often referred to as "sticky keys", and has been used by adversaries for unauthenticated access through a remote desktop login screen. [1]

Depending on the version of Windows, an adversary may take advantage of these features in different ways. Common methods used by adversaries include replacing accessibility feature binaries or pointers/references to these binaries in the Registry. In newer versions of Windows, the replaced binary needs to be digitally signed for x64 systems, the binary must reside in %systemdir%\, and it must be protected by Windows File or Resource Protection (WFP/WRP). [2] The Image File Execution Options Injection debugger method was likely discovered as a potential workaround because it does not require the corresponding accessibility feature binary to be replaced.

For simple binary replacement on Windows XP and later as well as and Windows Server 2003/R2 and later, for example, the program (e.g., C:\Windows\System32\utilman.exe) may be replaced with "cmd.exe" (or another program that provides backdoor access). Subsequently, pressing the appropriate key combination at the login screen while sitting at the keyboard or when connected over Remote Desktop Protocol will cause the replaced file to be executed with SYSTEM privileges. [3]

Other accessibility features exist that may also be leveraged in a similar fashion: [2][4]

* On-Screen Keyboard: C:\Windows\System32\osk.exe * Magnifier: C:\Windows\System32\Magnify.exe * Narrator: C:\Windows\System32\Narrator.exe * Display Switcher: C:\Windows\System32\DisplaySwitch.exe * App Switcher: C:\Windows\System32\AtBroker.exe

Adversaries may interrupt availability of system and network resources by inhibiting access to accounts utilized by legitimate users. Accounts may be deleted, locked, or manipulated (ex: changed credentials, revoked permissions for SaaS platforms such as Sharepoint) to remove access to accounts.[1] Adversaries may also subsequently log off and/or perform a System Shutdown/Reboot to set malicious changes into place.[2][3]

In Windows, Net utility, Set-LocalUser and Set-ADAccountPassword PowerShell cmdlets may be used by adversaries to modify user accounts. Accounts could also be disabled by Group Policy. In Linux, the passwd utility may be used to change passwords. On ESXi servers, accounts can be removed or modified via esxcli (`system account set`, `system account remove`).

Adversaries who use ransomware or similar attacks may first perform this and other Impact behaviors, such as Data Destruction and Defacement, in order to impede incident response/recovery before completing the Data Encrypted for Impact objective.

Adversaries may attempt to get a listing of valid accounts, usernames, or email addresses on a system or within a compromised environment. This information can help adversaries determine which accounts exist, which can aid in follow-on behavior such as brute-forcing, spear-phishing attacks, or account takeovers (e.g., Valid Accounts).

Adversaries may use several methods to enumerate accounts, including abuse of existing tools, built-in commands, and potential misconfigurations that leak account names and roles or permissions in the targeted environment.

For examples, cloud environments typically provide easily accessible interfaces to obtain user lists.[1][2] On hosts, adversaries can use default PowerShell and other command line functionality to identify accounts. Information about email addresses and accounts may also be extracted by searching an infected system’s files.

Adversaries may manipulate accounts to maintain and/or elevate access to victim systems. Account manipulation may consist of any action that preserves or modifies adversary access to a compromised account, such as modifying credentials or permission groups.[1] These actions could also include account activity designed to subvert security policies, such as performing iterative password updates to bypass password duration policies and preserve the life of compromised credentials.

In order to create or manipulate accounts, the adversary must already have sufficient permissions on systems or the domain. However, account manipulation may also lead to privilege escalation where modifications grant access to additional roles, permissions, or higher-privileged Valid Accounts.

Adversaries may activate firmware update mode on devices to prevent expected response functions from engaging in reaction to an emergency or process malfunction. For example, devices such as protection relays may have an operation mode designed for firmware installation. This mode may halt process monitoring and related functions to allow new firmware to be loaded. A device left in update mode may be placed in an inactive holding state if no firmware is provided to it. By entering and leaving a device in this mode, the adversary may deny its usual functionalities.

Adversaries may achieve persistence by adding a Registry key to the Active Setup of the local machine. Active Setup is a Windows mechanism that is used to execute programs when a user logs in. The value stored in the Registry key will be executed after a user logs into the computer.[1] These programs will be executed under the context of the user and will have the account's associated permissions level.

Adversaries may abuse Active Setup by creating a key under HKLM\SOFTWARE\Microsoft\Active Setup\Installed Components\ and setting a malicious value for StubPath. This value will serve as the program that will be executed when a user logs into the computer.[2][3][4][5][6]

Adversaries can abuse these components to execute malware, such as remote access tools, to maintain persistence through system reboots. Adversaries may also use Masquerading to make the Registry entries look as if they are associated with legitimate programs.

Adversaries may abuse Microsoft Office add-ins to obtain persistence on a compromised system. Office add-ins can be used to add functionality to Office programs. [1] There are different types of add-ins that can be used by the various Office products; including Word/Excel add-in Libraries (WLL/XLL), VBA add-ins, Office Component Object Model (COM) add-ins, automation add-ins, VBA Editor (VBE), Visual Studio Tools for Office (VSTO) add-ins, and Outlook add-ins. [2][3]

Add-ins can be used to obtain persistence because they can be set to execute code when an Office application starts.

Adversaries may grant additional permission levels to maintain persistent access to an adversary-controlled email account.

For example, the Add-MailboxPermission PowerShell cmdlet, available in on-premises Exchange and in the cloud-based service Office 365, adds permissions to a mailbox.[1][2][3] In Google Workspace, delegation can be enabled via the Google Admin console and users can delegate accounts via their Gmail settings.[4][5]

Adversaries may also assign mailbox folder permissions through individual folder permissions or roles. In Office 365 environments, adversaries may assign the Default or Anonymous user permissions or roles to the Top of Information Store (root), Inbox, or other mailbox folders. By assigning one or both user permissions to a folder, the adversary can utilize any other account in the tenant to maintain persistence to the target user’s mail folders.[6]

This may be used in persistent threat incidents as well as BEC (Business Email Compromise) incidents where an adversary can add Additional Cloud Roles to the accounts they wish to compromise. This may further enable use of additional techniques for gaining access to systems. For example, compromised business accounts are often used to send messages to other accounts in the network of the target business while creating inbox rules (ex: Internal Spearphishing), so the messages evade spam/phishing detection mechanisms.[7]

An adversary may add additional local or domain groups to an adversary-controlled account to maintain persistent access to a system or domain.

On Windows, accounts may use the `net localgroup` and `net group` commands to add existing users to local and domain groups.[1][2] On Linux, adversaries may use the `usermod` command for the same purpose.[3]

For example, accounts may be added to the local administrators group on Windows devices to maintain elevated privileges. They may also be added to the Remote Desktop Users group, which allows them to leverage Remote Desktop Protocol to log into the endpoints in the future.[4] Adversaries may also add accounts to VPN user groups to gain future persistence on the network.[5] On Linux, accounts may be added to the sudoers group, allowing them to persistently leverage Sudo and Sudo Caching for elevated privileges.

In Windows environments, machine accounts may also be added to domain groups. This allows the local SYSTEM account to gain privileges on the domain.[6]

Adversaries may attempt to position themselves between two or more networked devices using an adversary-in-the-middle (AiTM) technique to support follow-on behaviors such as Network Sniffing, Transmitted Data Manipulation, or replay attacks (Exploitation for Credential Access). By abusing features of common networking protocols that can determine the flow of network traffic (e.g. ARP, DNS, LLMNR, etc.), adversaries may force a device to communicate through an adversary controlled system so they can collect information or perform additional actions.[1]

For example, adversaries may manipulate victim DNS settings to enable other malicious activities such as preventing/redirecting users from accessing legitimate sites and/or pushing additional malware.[2][3][4] Adversaries may also manipulate DNS and leverage their position in order to intercept user credentials, including access tokens (Steal Application Access Token) and session cookies (Steal Web Session Cookie).[5][6] Downgrade Attacks can also be used to establish an AiTM position, such as by negotiating a less secure, deprecated, or weaker version of communication protocol (SSL/TLS) or encryption algorithm.[7][8][9]

Adversaries may also leverage the AiTM position to attempt to monitor and/or modify traffic, such as in Transmitted Data Manipulation. Adversaries can setup a position similar to AiTM to prevent traffic from flowing to the appropriate destination, potentially to impair defenses and/or in support of a Network Denial of Service.

Adversaries with privileged network access may seek to modify network traffic in real time using adversary-in-the-middle (AiTM) attacks. [1] This type of attack allows the adversary to intercept traffic to and/or from a particular device on the network. If a AiTM attack is established, then the adversary has the ability to block, log, modify, or inject traffic into the communication stream. There are several ways to accomplish this attack, but some of the most-common are Address Resolution Protocol (ARP) poisoning and the use of a proxy. [2]

An AiTM attack may allow an adversary to perform the following attacks: Block Reporting Message, Spoof Reporting Message, Modify Parameter, Unauthorized Command Message

Adversaries may target protection function alarms to prevent them from notifying operators of critical conditions. Alarm messages may be a part of an overall reporting system and of particular interest for adversaries. Disruption of the alarm system does not imply the disruption of the reporting system as a whole.

A Secura presentation on targeting OT notes a dual fold goal for adversaries attempting alarm suppression: prevent outgoing alarms from being raised and prevent incoming alarms from being responded to. [1] The method of suppression may greatly depend on the type of alarm in question:

* An alarm raised by a protocol message * An alarm signaled with I/O * An alarm bit set in a flag (and read)

In ICS environments, the adversary may have to suppress or contend with multiple alarms and/or alarm propagation to achieve a specific goal to evade detection or prevent intended responses from occurring. [1] Methods of suppression may involve tampering or altering device displays and logs, modifying in memory code to fixed values, or even tampering with assembly level instruction code.

Adversaries may establish persistence and/or elevate privileges by executing malicious content triggered by AppCert DLLs loaded into processes. Dynamic-link libraries (DLLs) that are specified in the AppCertDLLs Registry key under HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Session Manager\ are loaded into every process that calls the ubiquitously used application programming interface (API) functions CreateProcess, CreateProcessAsUser, CreateProcessWithLoginW, CreateProcessWithTokenW, or WinExec. [1]

Similar to Process Injection, this value can be abused to obtain elevated privileges by causing a malicious DLL to be loaded and run in the context of separate processes on the computer. Malicious AppCert DLLs may also provide persistence by continuously being triggered by API activity.

Adversaries may execute their own malicious payloads by hijacking how the .NET `AppDomainManager` loads assemblies. The .NET framework uses the `AppDomainManager` class to create and manage one or more isolated runtime environments (called application domains) inside a process to host the execution of .NET applications. Assemblies (`.exe` or `.dll` binaries compiled to run as .NET code) may be loaded into an application domain as executable code.[1]

Known as "AppDomainManager injection," adversaries may execute arbitrary code by hijacking how .NET applications load assemblies. For example, malware may create a custom application domain inside a target process to load and execute an arbitrary assembly. Alternatively, configuration files (`.config`) or process environment variables that define .NET runtime settings may be tampered with to instruct otherwise benign .NET applications to load a malicious assembly (identified by name) into the target process.[2][3][4]

Adversaries may establish persistence and/or elevate privileges by executing malicious content triggered by AppInit DLLs loaded into processes. Dynamic-link libraries (DLLs) that are specified in the AppInit_DLLs value in the Registry keys HKEY_LOCAL_MACHINE\Software\Microsoft\Windows NT\CurrentVersion\Windows or HKEY_LOCAL_MACHINE\Software\Wow6432Node\Microsoft\Windows NT\CurrentVersion\Windows are loaded by user32.dll into every process that loads user32.dll. In practice this is nearly every program, since user32.dll is a very common library. [1]

Similar to Process Injection, these values can be abused to obtain elevated privileges by causing a malicious DLL to be loaded and run in the context of separate processes on the computer. [2] Malicious AppInit DLLs may also provide persistence by continuously being triggered by API activity.

The AppInit DLL functionality is disabled in Windows 8 and later versions when secure boot is enabled. [3]

Adversaries may target resource intensive features of applications to cause a denial of service (DoS), denying availability to those applications. For example, specific features in web applications may be highly resource intensive. Repeated requests to those features may be able to exhaust system resources and deny access to the application or the server itself.[1]

Adversaries may communicate using OSI application layer protocols to avoid detection/network filtering by blending in with existing traffic. Commands to the remote system, and often the results of those commands, will be embedded within the protocol traffic between the client and server.

Adversaries may utilize many different protocols, including those used for web browsing, transferring files, electronic mail, DNS, or publishing/subscribing. For connections that occur internally within an enclave (such as those between a proxy or pivot node and other nodes), commonly used protocols are SMB, SSH, or RDP.[1]

Adversaries may establish persistence and/or elevate privileges by executing malicious content triggered by application shims. The Microsoft Windows Application Compatibility Infrastructure/Framework (Application Shim) was created to allow for backward compatibility of software as the operating system codebase changes over time. For example, the application shimming feature allows developers to apply fixes to applications (without rewriting code) that were created for Windows XP so that it will work with Windows 10. [1]

Within the framework, shims are created to act as a buffer between the program (or more specifically, the Import Address Table) and the Windows OS. When a program is executed, the shim cache is referenced to determine if the program requires the use of the shim database (.sdb). If so, the shim database uses hooking to redirect the code as necessary in order to communicate with the OS.

A list of all shims currently installed by the default Windows installer (sdbinst.exe) is kept in:

* %WINDIR%\AppPatch\sysmain.sdb and * hklm\software\microsoft\windows nt\currentversion\appcompatflags\installedsdb

Custom databases are stored in:

* %WINDIR%\AppPatch\custom & %WINDIR%\AppPatch\AppPatch64\Custom and * hklm\software\microsoft\windows nt\currentversion\appcompatflags\custom

To keep shims secure, Windows designed them to run in user mode so they cannot modify the kernel and you must have administrator privileges to install a shim. However, certain shims can be used to Bypass User Account Control (UAC and RedirectEXE), inject DLLs into processes (InjectDLL), disable Data Execution Prevention (DisableNX) and Structure Exception Handling (DisableSEH), and intercept memory addresses (GetProcAddress).

Utilizing these shims may allow an adversary to perform several malicious acts such as elevate privileges, install backdoors, disable defenses like Windows Defender, etc. [2] Shims can also be abused to establish persistence by continuously being invoked by affected programs.

Adversaries may attempt to get a listing of open application windows. Window listings could convey information about how the system is used.[1] For example, information about application windows could be used identify potential data to collect as well as identifying security tooling (Security Software Discovery) to evade.[2]

Adversaries typically abuse system features for this type of enumeration. For example, they may gather information through native system features such as Command and Scripting Interpreter commands and Native API functions.

Adversaries may exploit software vulnerabilities that can cause an application or system to crash and deny availability to users. [1] Some systems may automatically restart critical applications and services when crashes occur, but they can likely be re-exploited to cause a persistent denial of service (DoS) condition.

Adversaries may exploit known or zero-day vulnerabilities to crash applications and/or systems, which may also lead to dependent applications and/or systems to be in a DoS condition. Crashed or restarted applications or systems may also have other effects such as Data Destruction, Firmware Corruption, Service Stop etc. which may further cause a DoS condition and deny availability to critical information, applications and/or systems.

An adversary may compress and/or encrypt data that is collected prior to exfiltration. Compressing the data can help to obfuscate the collected data and minimize the amount of data sent over the network.[1] Encryption can be used to hide information that is being exfiltrated from detection or make exfiltration less conspicuous upon inspection by a defender.

Both compression and encryption are done prior to exfiltration, and can be performed using a utility, 3rd party library, or custom method.

An adversary may compress or encrypt data that is collected prior to exfiltration using a custom method. Adversaries may choose to use custom archival methods, such as encryption with XOR or stream ciphers implemented with no external library or utility references. Custom implementations of well-known compression algorithms have also been used.[1]

An adversary may compress or encrypt data that is collected prior to exfiltration using 3rd party libraries. Many libraries exist that can archive data, including Python rarfile [1], libzip [2], and zlib [3]. Most libraries include functionality to encrypt and/or compress data.

Some archival libraries are preinstalled on systems, such as bzip2 on macOS and Linux, and zip on Windows. Note that the libraries are different from the utilities. The libraries can be linked against when compiling, while the utilities require spawning a subshell, or a similar execution mechanism.

Adversaries may use utilities to compress and/or encrypt collected data prior to exfiltration. Many utilities include functionalities to compress, encrypt, or otherwise package data into a format that is easier/more secure to transport.

Adversaries may abuse various utilities to compress or encrypt data before exfiltration. Some third party utilities may be preinstalled, such as tar on Linux and macOS or zip on Windows systems.

On Windows, diantz or makecab may be used to package collected files into a cabinet (.cab) file. diantz may also be used to download and compress files from remote locations (i.e. Remote Data Staging).[1] xcopy on Windows can copy files and directories with a variety of options. Additionally, adversaries may use certutil to Base64 encode collected data before exfiltration.

Adversaries may use also third party utilities, such as 7-Zip, WinRAR, and WinZip, to perform similar activities.[2][3][4]

Adversaries may employ a known asymmetric encryption algorithm to conceal command and control traffic rather than relying on any inherent protections provided by a communication protocol. Asymmetric cryptography, also known as public key cryptography, uses a keypair per party: one public that can be freely distributed, and one private. Due to how the keys are generated, the sender encrypts data with the receiver’s public key and the receiver decrypts the data with their private key. This ensures that only the intended recipient can read the encrypted data. Common public key encryption algorithms include RSA and ElGamal.

For efficiency, many protocols (including SSL/TLS) use symmetric cryptography once a connection is established, but use asymmetric cryptography to establish or transmit a key. As such, these protocols are classified as Asymmetric Cryptography.[1][2]

Adversaries may inject malicious code into processes via the asynchronous procedure call (APC) queue in order to evade process-based defenses as well as possibly elevate privileges. APC injection is a method of executing arbitrary code in the address space of a separate live process.

APC injection is commonly performed by attaching malicious code to the APC Queue [1] of a process's thread. Queued APC functions are executed when the thread enters an alterable state.[1] A handle to an existing victim process is first created with native Windows API calls such as OpenThread. At this point QueueUserAPC can be used to invoke a function (such as LoadLibrayA pointing to a malicious DLL).

A variation of APC injection, dubbed "Early Bird injection", involves creating a suspended process in which malicious code can be written and executed before the process' entry point (and potentially subsequent anti-malware hooks) via an APC. [2] AtomBombing [3] is another variation that utilizes APCs to invoke malicious code previously written to the global atom table.[4]

Running code in the context of another process may allow access to the process's memory, system/network resources, and possibly elevated privileges. Execution via APC injection may also evade detection from security products since the execution is masked under a legitimate process.

Adversaries may abuse the at utility to perform task scheduling for initial or recurring execution of malicious code. The at utility exists as an executable within Windows, Linux, and macOS for scheduling tasks at a specified time and date. Although deprecated in favor of Scheduled Task's schtasks in Windows environments, using at requires that the Task Scheduler service be running, and the user to be logged on as a member of the local Administrators group. In addition to explicitly running the `at` command, adversaries may also schedule a task with at by directly leveraging the Windows Management Instrumentation `Win32_ScheduledJob` WMI class.[1]

On Linux and macOS, at may be invoked by the superuser as well as any users added to the at.allow file. If the at.allow file does not exist, the at.deny file is checked. Every username not listed in at.deny is allowed to invoke at. If the at.deny exists and is empty, global use of at is permitted. If neither file exists (which is often the baseline) only the superuser is allowed to use at.[2]

Adversaries may use at to execute programs at system startup or on a scheduled basis for Persistence. at can also be abused to conduct remote Execution as part of Lateral Movement and/or to run a process under the context of a specified account (such as SYSTEM).

In Linux environments, adversaries may also abuse at to break out of restricted environments by using a task to spawn an interactive system shell or to run system commands. Similarly, at may also be used for Privilege Escalation if the binary is allowed to run as superuser via sudo.[3]

An adversary can leverage a computer's peripheral devices (e.g., microphones and webcams) or applications (e.g., voice and video call services) to capture audio recordings for the purpose of listening into sensitive conversations to gather information.[1]

Malware or scripts may be used to interact with the devices through an available API provided by the operating system or an application to capture audio. Audio files may be written to disk and exfiltrated later.

Adversaries may abuse authentication packages to execute DLLs when the system boots. Windows authentication package DLLs are loaded by the Local Security Authority (LSA) process at system start. They provide support for multiple logon processes and multiple security protocols to the operating system.[1]

Adversaries can use the autostart mechanism provided by LSA authentication packages for persistence by placing a reference to a binary in the Windows Registry location HKLM\SYSTEM\CurrentControlSet\Control\Lsa\ with the key value of "Authentication Packages"=<target binary>. The binary will then be executed by the system when the authentication packages are loaded.

Adversaries may execute commands and perform malicious tasks using AutoIT and AutoHotKey automation scripts. AutoIT and AutoHotkey (AHK) are scripting languages that enable users to automate Windows tasks. These automation scripts can be used to perform a wide variety of actions, such as clicking on buttons, entering text, and opening and closing programs.[1][2]

Adversaries may use AHK (`.ahk`) and AutoIT (`.au3`) scripts to execute malicious code on a victim's system. For example, adversaries have used for AHK to execute payloads and other modular malware such as keyloggers. Adversaries have also used custom AHK files containing embedded malware as Phishing payloads.[3]

These scripts may also be compiled into self-contained executable payloads (`.exe`).[1][2]

Once established within a system or network, an adversary may use automated techniques for collecting internal data. Methods for performing this technique could include use of a Command and Scripting Interpreter to search for and copy information fitting set criteria such as file type, location, or name at specific time intervals.

In cloud-based environments, adversaries may also use cloud APIs, data pipelines, command line interfaces, or extract, transform, and load (ETL) services to automatically collect data.[1]

This functionality could also be built into remote access tools.

This technique may incorporate use of other techniques such as File and Directory Discovery and Lateral Tool Transfer to identify and move files, as well as Cloud Service Dashboard and Cloud Storage Object Discovery to identify resources in cloud environments.

Adversaries may automate collection of industrial environment information using tools or scripts. This automated collection may leverage native control protocols and tools available in the control systems environment. For example, the OPC protocol may be used to enumerate and gather information. Access to a system or interface with these native protocols may allow collection and enumeration of other attached, communicating servers and devices.

Adversaries may exfiltrate data, such as sensitive documents, through the use of automated processing after being gathered during Collection.[1]

When automated exfiltration is used, other exfiltration techniques likely apply as well to transfer the information out of the network, such as Exfiltration Over C2 Channel and Exfiltration Over Alternative Protocol.

Adversaries may leverage AutoRun functionality or scripts to execute malicious code. Devices configured to enable AutoRun functionality or legacy operating systems may be susceptible to abuse of these features to run malicious code stored on various forms of removeable media (i.e., USB, Disk Images [.ISO]). Commonly, AutoRun or AutoPlay are disabled in many operating systems configurations to mitigate against this technique. If a device is configured to enable AutoRun or AutoPlay, adversaries may execute code on the device by mounting the removable media to the device, either through physical or virtual means. This may be especially relevant for virtual machine environments where disk images may be dynamically mapped to a guest system on a hypervisor.

An example could include an adversary gaining access to a hypervisor through the management interface to modify a virtual machine’s hardware configuration. They could then deploy an iso image with a malicious AutoRun script to cause the virtual machine to automatically execute the code contained on the disk image. This would enable the execution of malicious code within a virtual machine without needing any prior remote access to that system.

Adversaries may abuse BITS jobs to persistently execute code and perform various background tasks. Windows Background Intelligent Transfer Service (BITS) is a low-bandwidth, asynchronous file transfer mechanism exposed through Component Object Model (COM).[1][2] BITS is commonly used by updaters, messengers, and other applications preferred to operate in the background (using available idle bandwidth) without interrupting other networked applications. File transfer tasks are implemented as BITS jobs, which contain a queue of one or more file operations.

The interface to create and manage BITS jobs is accessible through PowerShell and the BITSAdmin tool.[2][3]

Adversaries may abuse BITS to download (e.g. Ingress Tool Transfer), execute, and even clean up after running malicious code (e.g. Indicator Removal). BITS tasks are self-contained in the BITS job database, without new files or registry modifications, and often permitted by host firewalls.[4][5][6] BITS enabled execution may also enable persistence by creating long-standing jobs (the default maximum lifetime is 90 days and extendable) or invoking an arbitrary program when a job completes or errors (including after system reboots).[7][4]

BITS upload functionalities can also be used to perform Exfiltration Over Alternative Protocol.[4]

Adversaries may attempt to get a listing of backup software or configurations that are installed on a system. Adversaries may use this information to shape follow-on behaviors, such as Data Destruction, Inhibit System Recovery, or Data Encrypted for Impact.

Commands that can be used to obtain security software information are netsh, `reg query` with Reg, `dir` with cmd, and Tasklist, but other indicators of discovery behavior may be more specific to the type of software or security system the adversary is looking for, such as Veeam, Acronis, Dropbox, or Paragon.[1]

Adversaries may leverage the network bandwidth resources of co-opted systems to complete resource-intensive tasks, which may impact system and/or hosted service availability.

Adversaries may also use malware that leverages a system's network bandwidth as part of a botnet in order to facilitate Network Denial of Service campaigns and/or to seed malicious torrents.[1] Alternatively, they may engage in proxyjacking by selling use of the victims' network bandwidth and IP address to proxyware services.[2] Finally, they may engage in internet-wide scanning in order to identify additional targets for compromise.[3]

In addition to incurring potential financial costs or availability disruptions, this technique may cause reputational damage if a victim’s bandwidth is used for illegal activities.[2]

Adversaries may use an existing, legitimate external Web service as a means for sending commands to and receiving output from a compromised system over the Web service channel. Compromised systems may leverage popular websites and social media to host command and control (C2) instructions. Those infected systems can then send the output from those commands back over that Web service channel. The return traffic may occur in a variety of ways, depending on the Web service being utilized. For example, the return traffic may take the form of the compromised system posting a comment on a forum, issuing a pull request to development project, updating a document hosted on a Web service, or by sending a Tweet.

Popular websites and social media acting as a mechanism for C2 may give a significant amount of cover due to the likelihood that hosts within a network are already communicating with them prior to a compromise. Using common services, such as those offered by Google or Twitter, makes it easier for adversaries to hide in expected noise. Web service providers commonly use SSL/TLS encryption, giving adversaries an added level of protection.

Adversaries may use binary padding to add junk data and change the on-disk representation of malware. This can be done without affecting the functionality or behavior of a binary, but can increase the size of the binary beyond what some security tools are capable of handling due to file size limitations.

Binary padding effectively changes the checksum of the file and can also be used to avoid hash-based blocklists and static anti-virus signatures.[1] The padding used is commonly generated by a function to create junk data and then appended to the end or applied to sections of malware.[2] Increasing the file size may decrease the effectiveness of certain tools and detection capabilities that are not designed or configured to scan large files. This may also reduce the likelihood of being collected for analysis. Public file scanning services, such as VirusTotal, limits the maximum size of an uploaded file to be analyzed.[3]

Operational technology communications occur over serial COM, Ethernet, Wi-Fi, cellular (4G/5G), and satellite mediums. Adversaries may block communications to prevent reporting messages and command messages from reaching their intended target devices disrupting processes, operations, and causing cyber-physical impacts.[1]

Adversaries may block communications by either making modifications to software (System Firmware, Module Firmware, Hooking, and Rootkit) and services (Service Stop, Denial of Service) on systems and devices or by positioning themselves between systems and devices and intercepting and blocking the communications such as the case with an Adversary-in-the-Middle attack.

Adversaries may block messages between systems and devices in an OT/ICS environment to disrupt processes. Messages typically fall into two categories: (1) reporting messages that contain telemetry data about the current state of systems, devices, and processes and (2) command messages that contain instructions to control systems, devices, and processes. Both types of messages are critical for the proper functioning of industrial control processes and failure of the messages to reach their intended destinations could inhibit response functions or create an unsafe condition that could have physical impacts.[1][2]

Adversaries may block communications by either making modifications to software (System Firmware, Module Firmware, Hooking, and Rootkit) and services (Service Stop, Denial of Service) on systems and devices or by positioning themselves between systems and devices and intercepting and blocking the communications such as the case with an Adversary-in-the-Middle attack.

Adversaries may configure system settings to automatically execute a program during system boot or logon to maintain persistence or gain higher-level privileges on compromised systems. Operating systems may have mechanisms for automatically running a program on system boot or account logon.[1][2][3][4][5] These mechanisms may include automatically executing programs that are placed in specially designated directories or are referenced by repositories that store configuration information, such as the Windows Registry. An adversary may achieve the same goal by modifying or extending features of the kernel.

Since some boot or logon autostart programs run with higher privileges, an adversary may leverage these to elevate privileges.

Adversaries may use scripts automatically executed at boot or logon initialization to establish persistence.[1][2] Initialization scripts can be used to perform administrative functions, which may often execute other programs or send information to an internal logging server. These scripts can vary based on operating system and whether applied locally or remotely.

Adversaries may use these scripts to maintain persistence on a single system. Depending on the access configuration of the logon scripts, either local credentials or an administrator account may be necessary.

An adversary may also be able to escalate their privileges since some boot or logon initialization scripts run with higher privileges.

Adversaries may use bootkits to persist on systems. A bootkit is a malware variant that modifies the boot sectors of a hard drive, allowing malicious code to execute before a computer's operating system has loaded. Bootkits reside at a layer below the operating system and may make it difficult to perform full remediation unless an organization suspects one was used and can act accordingly.

In BIOS systems, a bootkit may modify the Master Boot Record (MBR) and/or Volume Boot Record (VBR).[1] The MBR is the section of disk that is first loaded after completing hardware initialization by the BIOS. It is the location of the boot loader. An adversary who has raw access to the boot drive may overwrite this area, diverting execution during startup from the normal boot loader to adversary code.[2]

The MBR passes control of the boot process to the VBR. Similar to the case of MBR, an adversary who has raw access to the boot drive may overwrite the VBR to divert execution during startup to adversary code.

In UEFI (Unified Extensible Firmware Interface) systems, a bootkit may instead create or modify files in the EFI system partition (ESP). The ESP is a partition on data storage used by devices containing UEFI that allows the system to boot the OS and other utilities used by the system. An adversary can use the newly created or patched files in the ESP to run malicious kernel code.[3][4]

Adversaries may perform broadcast discovery requests to enumerate systems and devices on a network. Broadcast discovery works by one system or device sending messages to all systems and devices on a network (or subnet) and then waiting for a response. If a response is received that means the system or device that responded is live and can communicate over that protocol. Adversaries may leverage different protocols supported on the network for sending broadcast messages.

Some common OT protocols that have broadcast discovery mechanisms are Building Automation and Control Network (BACNet) Who-Is requests, Common Industrial Protocol (CIP) List Identity User Datagram Protocol (UDP) broadcast requests, and Siemens S7 broadcast identification requests.[1][2]

Adversaries may abuse internet browser extensions to establish persistent access to victim systems. Browser extensions or plugins are small programs that can add functionality to and customize aspects of internet browsers. They can be installed directly via a local file or custom URL or through a browser's app store - an official online platform where users can browse, install, and manage extensions for a specific web browser. Extensions generally inherit the web browser's permissions previously granted.[1][2] Malicious extensions can be installed into a browser through malicious app store downloads masquerading as legitimate extensions, through social engineering, or by an adversary that has already compromised a system. Security can be limited on browser app stores, so it may not be difficult for malicious extensions to defeat automated scanners.[3] Depending on the browser, adversaries may also manipulate an extension's update url to install updates from an adversary-controlled server or manipulate the mobile configuration file to silently install additional extensions.

Adversaries may abuse how chromium-based browsers load extensions by modifying or replacing the Preferences and/or Secure Preferences files to silently install malicious extensions. When the browser is not running, adversaries can alter these files, ensuring the extension is loaded, granted desired permissions, and will persist in browser sessions. This method does not require user consent and extensions are silently loaded in the background from disk or from the browser's trusted store.[4] Previous to macOS 11, adversaries could silently install browser extensions via the command line using the profiles tool to install malicious .mobileconfig files. In macOS 11+, the use of the profiles tool can no longer install configuration profiles; however, .mobileconfig files can be planted and installed with user interaction.[5] Once the extension is installed, it can browse to websites in the background, steal all information that a user enters into a browser (including credentials), and be used as an installer for a RAT for persistence.[6][7][8][9]

There have also been instances of botnets using a persistent backdoor through malicious Chrome extensions for Command and Control.[10][11] Adversaries may also use browser extensions to modify browser permissions and components, privacy settings, and other security controls for Stealth.[12][13]

Adversaries may attempt to blend in with legitimate traffic by spoofing browser and system attributes like operating system, system language, platform, user-agent string, resolution, time zone, etc. The HTTP User-Agent request header is a string that lets servers and network peers identify the application, operating system, vendor, and/or version of the requesting user agent.[1]

Adversaries may gather this information through System Information Discovery or by users navigating to adversary-controlled websites, and then use that information to craft their web traffic to evade defenses.[2]

Adversaries may enumerate information about browsers to learn more about compromised environments. Data saved by browsers (such as bookmarks, accounts, and browsing history) may reveal a variety of personal information about users (e.g., banking sites, relationships/interests, social media, etc.) as well as details about internal network resources such as servers, tools/dashboards, or other related infrastructure.[1]

Browser information may also highlight additional targets after an adversary has access to valid credentials, especially Credentials In Files associated with logins cached by a browser.

Specific storage locations vary based on platform and/or application, but browser information is typically stored in local files and databases (e.g., `%APPDATA%/Google/Chrome`).[2]

Adversaries may take advantage of security vulnerabilities and inherent functionality in browser software to change content, modify user-behaviors, and intercept information as part of various browser session hijacking techniques.[1]

A specific example is when an adversary injects software into a browser that allows them to inherit cookies, HTTP sessions, and SSL client certificates of a user then use the browser as a way to pivot into an authenticated intranet.[2][3] Executing browser-based behaviors such as pivoting may require specific process permissions, such as SeDebugPrivilege and/or high-integrity/administrator rights.

Another example involves pivoting browser traffic from the adversary's browser through the user's browser by setting up a proxy which will redirect web traffic. This does not alter the user's traffic in any way, and the proxy connection can be severed as soon as the browser is closed. The adversary assumes the security context of whichever browser process the proxy is injected into. Browsers typically create a new process for each tab that is opened and permissions and certificates are separated accordingly. With these permissions, an adversary could potentially browse to any resource on an intranet, such as Sharepoint or webmail, that is accessible through the browser and which the browser has sufficient permissions. Browser pivoting may also bypass security provided by 2-factor authentication.[4]

Adversaries may use brute force techniques to gain access to accounts when passwords are unknown or when password hashes are obtained.[1] Without knowledge of the password for an account or set of accounts, an adversary may systematically guess the password using a repetitive or iterative mechanism.[2] Brute forcing passwords can take place via interaction with a service that will check the validity of those credentials or offline against previously acquired credential data, such as password hashes.

Brute forcing credentials may take place at various points during a breach. For example, adversaries may attempt to brute force access to Valid Accounts within a victim environment leveraging knowledge gathered from other post-compromise behaviors such as OS Credential Dumping, Account Discovery, or Password Policy Discovery. Adversaries may also combine brute forcing activity with behaviors such as External Remote Services as part of Initial Access.

If an adversary guesses the correct password but fails to login to a compromised account due to location-based conditional access policies, they may change their infrastructure until they match the victim’s location and therefore bypass those policies.[3]

Adversaries may repetitively or successively change I/O point values to perform an action. Brute Force I/O may be achieved by changing either a range of I/O point values or a single point value repeatedly to manipulate a process function. The adversary's goal and the information they have about the target environment will influence which of the options they choose. In the case of brute forcing a range of point values, the adversary may be able to achieve an impact without targeting a specific point. In the case where a single point is targeted, the adversary may be able to generate instability on the process function associated with that particular point.

Adversaries may use Brute Force I/O to cause failures within various industrial processes. These failures could be the result of wear on equipment or damage to downstream equipment.

Adversaries may bypass UAC mechanisms to elevate process privileges on system. Windows User Account Control (UAC) allows a program to elevate its privileges (tracked as integrity levels ranging from low to high) to perform a task under administrator-level permissions, possibly by prompting the user for confirmation. The impact to the user ranges from denying the operation under high enforcement to allowing the user to perform the action if they are in the local administrators group and click through the prompt or allowing them to enter an administrator password to complete the action.[1]

If the UAC protection level of a computer is set to anything but the highest level, certain Windows programs can elevate privileges or execute some elevated Component Object Model objects without prompting the user through the UAC notification box.[2][3] An example of this is use of Rundll32 to load a specifically crafted DLL which loads an auto-elevated Component Object Model object and performs a file operation in a protected directory which would typically require elevated access. Malicious software may also be injected into a trusted process to gain elevated privileges without prompting a user.[4]

Many methods have been discovered to bypass UAC. The Github readme page for UACME contains an extensive list of methods[5] that have been discovered and implemented, but may not be a comprehensive list of bypasses. Additional bypass methods are regularly discovered and some used in the wild, such as:

* eventvwr.exe can auto-elevate and execute a specified binary or script.[6][7]

Another bypass is possible through some lateral movement techniques if credentials for an account with administrator privileges are known, since UAC is a single system security mechanism, and the privilege or integrity of a process running on one system will be unknown on remote systems and default to high integrity.[8]

Adversaries may abuse CMSTP to proxy execution of malicious code. The Microsoft Connection Manager Profile Installer (CMSTP.exe) is a command-line program used to install Connection Manager service profiles. [1] CMSTP.exe accepts an installation information file (INF) as a parameter and installs a service profile leveraged for remote access connections.

Adversaries may supply CMSTP.exe with INF files infected with malicious commands. [2] Similar to Regsvr32 / ”Squiblydoo”, CMSTP.exe may be abused to load and execute DLLs [3] and/or COM scriptlets (SCT) from remote servers. [4] [5] [6] This execution may also bypass AppLocker and other application control defenses since CMSTP.exe is a legitimate binary that may be signed by Microsoft.

CMSTP.exe can also be abused to Bypass User Account Control and execute arbitrary commands from a malicious INF through an auto-elevated COM interface. [3] [5] [6]

Adversaries may leverage the COR_PROFILER environment variable to hijack the execution flow of programs that load the .NET CLR. The COR_PROFILER is a .NET Framework feature which allows developers to specify an unmanaged (or external of .NET) profiling DLL to be loaded into each .NET process that loads the Common Language Runtime (CLR). These profilers are designed to monitor, troubleshoot, and debug managed code executed by the .NET CLR.[1][2]

The COR_PROFILER environment variable can be set at various scopes (system, user, or process) resulting in different levels of influence. System and user-wide environment variable scopes are specified in the Registry, where a Component Object Model (COM) object can be registered as a profiler DLL. A process scope COR_PROFILER can also be created in-memory without modifying the Registry. Starting with .NET Framework 4, the profiling DLL does not need to be registered as long as the location of the DLL is specified in the COR_PROFILER_PATH environment variable.[2]

Adversaries may abuse COR_PROFILER to establish persistence that executes a malicious DLL in the context of all .NET processes every time the CLR is invoked. The COR_PROFILER can also be used to elevate privileges (ex: Bypass User Account Control) if the victim .NET process executes at a higher permission level, as well as to hook and impair defenses provided by .NET processes.[3][4][5][6][7]

Adversaries may attempt to access cached domain credentials used to allow authentication to occur in the event a domain controller is unavailable.[1]

On Windows Vista and newer, the hash format is DCC2 (Domain Cached Credentials version 2) hash, also known as MS-Cache v2 hash.[2] The number of default cached credentials varies and can be altered per system. This hash does not allow pass-the-hash style attacks, and instead requires Password Cracking to recover the plaintext password.[3]

On Linux systems, Active Directory credentials can be accessed through caches maintained by software like System Security Services Daemon (SSSD) or Quest Authentication Services (formerly VAS). Cached credential hashes are typically located at `/var/lib/sss/db/cache.[domain].ldb` for SSSD or `/var/opt/quest/vas/authcache/vas_auth.vdb` for Quest. Adversaries can use utilities, such as `tdbdump`, on these database files to dump the cached hashes and use Password Cracking to obtain the plaintext password.[4]

With SYSTEM or sudo access, the tools/utilities such as Mimikatz, Reg, and secretsdump.py for Windows or Linikatz for Linux can be used to extract the cached credentials.[4]

Note: Cached credentials for Windows Vista are derived using PBKDF2.[2]