ATT&CK Techniques

Overview

Techniques describe "how" an adversary achieves a tactical objective. While tactics answer "why" (e.g., the adversary wants to escalate privileges), techniques answer "how" (e.g., the adversary exploits a SUID binary). Each technique belongs to one or more tactics and includes a description, procedure examples from real-world incidents, detection guidance, and mitigations.

Techniques are identified by T-numbers (e.g., T1059 — Command and Scripting Interpreter). Sub-techniques provide more specific variants (e.g., T1059.001 — PowerShell). Procedures are the specific implementations observed in the wild — how a particular threat group used a technique in an actual operation.

Key Concepts

Technique Structure

Every ATT&CK technique page contains:

Section What It Tells You
Description What the technique is and how it works
Sub-Techniques More specific variants (when applicable)
Procedure Examples Real-world usage by threat groups and malware
Mitigations Defensive measures that reduce risk
Detection How to identify this technique in telemetry
References Source reports and research

Techniques vs Sub-Techniques

Sub-techniques break down a broad technique into specific variants. The parent technique describes the general approach; sub-techniques describe specific implementations.

Example — T1059 (Command and Scripting Interpreter):

ID Name Description
T1059 Command and Scripting Interpreter Adversary uses command-line interfaces or scripting to execute commands
T1059.001 PowerShell Execution via PowerShell
T1059.003 Windows Command Shell Execution via cmd.exe
T1059.004 Unix Shell Execution via Bash, sh, or other Unix shells
T1059.005 Visual Basic Execution via VBScript or VBA macros
T1059.006 Python Execution via Python interpreter

Not every technique has sub-techniques. Some techniques are specific enough on their own (e.g., T1190 — Exploit Public-Facing Application).

Techniques Across Multiple Tactics

Some techniques appear under more than one tactic because the same action can serve different objectives. For example:

Technique Tactics
T1053 — Scheduled Task/Job Execution, Persistence, Privilege Escalation
T1078 — Valid Accounts Initial Access, Persistence, Privilege Escalation, Defense Evasion
T1055 — Process Injection Defense Evasion, Privilege Escalation

T1078 (Valid Accounts) maps to four tactics because stolen credentials serve multiple purposes — gaining initial access, persisting in the environment, escalating privileges, and blending in with legitimate activity.

Procedure Examples

Procedures are the specific, observed implementations of a technique. They tie abstract techniques to real adversary behavior.

For example, under T1003 (OS Credential Dumping): - APT28 used Mimikatz to dump credentials from LSASS memory

Procedure examples make techniques concrete. They show how real attackers actually use a technique, not just how it could theoretically be used.

Key Technique Categories

The following are some of the most commonly encountered technique areas in penetration testing. These are representative examples — not an exhaustive list.

Initial Access Techniques:

ID Name Common Usage
T1190 Exploit Public-Facing Application Web app exploitation (SQLi, RCE)
T1566 Phishing Spearphishing with attachments, links, or via services
T1078 Valid Accounts Using stolen or default credentials
T1195 Supply Chain Compromise Compromising software update mechanisms

Execution Techniques:

ID Name Common Usage
T1059 Command and Scripting Interpreter Shell commands, PowerShell, Python
T1053 Scheduled Task/Job Cron jobs, Windows Task Scheduler
T1047 Windows Management Instrumentation Remote execution via WMI

Privilege Escalation Techniques:

ID Name Common Usage
T1068 Exploitation for Privilege Escalation Kernel exploits, service exploits
T1548 Abuse Elevation Control Mechanism Sudo misconfig, UAC bypass
T1134 Access Token Manipulation Token impersonation on Windows

Credential Access Techniques:

ID Name Common Usage
T1003 OS Credential Dumping LSASS dump, SAM extraction, /etc/shadow
T1558 Steal or Forge Kerberos Tickets Kerberoasting, Golden Ticket
T1110 Brute Force Password spraying, credential stuffing

Lateral Movement Techniques:

ID Name Common Usage
T1021 Remote Services RDP, SSH, SMB, WinRM
T1550 Use Alternate Authentication Material Pass-the-hash, pass-the-ticket
T1570 Lateral Tool Transfer Moving tools between compromised hosts

Technique IDs in Practice

ATT&CK IDs follow a consistent format:

Format Example Meaning
TA00XX TA0001 Tactic (Initial Access)
T1XXX T1059 Technique (Command and Scripting Interpreter)
T1XXX.00X T1059.001 Sub-technique (PowerShell)
GXXXX G0007 Threat group (APT28)
SXXXX S0002 Software (Mimikatz)
MXXXX M1036 Mitigation (Account Use Policies)

Data Sources and Detection

Each technique lists data sources — the telemetry needed to detect it. Data sources tell defenders what to log and monitor.

Example — T1059.001 (PowerShell): - Process creation — monitor for powershell.exe launches - Command execution — log PowerShell script block logging (Event ID 4104) - Module loading — track suspicious module imports

Data sources help security teams answer: "Can we detect this technique with our current logging?" If the answer is no, the data source tells them what to enable.

Mitigations

Mitigations are preventive measures mapped to techniques. They answer: "How do we reduce the risk of this technique?"

Example mitigations for T1059 (Command and Scripting Interpreter): - M1042 — Disable or Remove Feature or Program: Remove unnecessary scripting interpreters - M1049 — Antivirus/Antimalware: Detect known malicious scripts - M1038 — Execution Prevention: Application whitelisting to block unauthorized interpreters

Practical Examples

Reading a Technique Page

When examining a technique page on attack.mitre.org:

  1. Read the description — understand what the technique does
  2. Check sub-techniques — determine if a more specific variant applies
  3. Review procedure examples — see how real adversaries used it
  4. Check detection guidance — identify what telemetry is needed
  5. Review mitigations — determine what preventive controls apply
  6. Note the data sources — verify your logging covers them

Mapping Pentest Findings to Techniques

Finding: Gained root via writable cron job
Mapping:
  Tactic: TA0004 (Privilege Escalation)
  Technique: T1053.003 (Scheduled Task/Job: Cron)

Finding: Extracted domain hashes via DCSync
Mapping:
  Tactic: TA0006 (Credential Access)
  Technique: T1003.006 (OS Credential Dumping: DCSync)

Finding: Moved to file server using stolen NTLM hash
Mapping:
  Tactic: TA0008 (Lateral Movement)
  Technique: T1550.002 (Use Alternate Authentication Material: Pass the Hash)

References

Official Documentation