Red TeamingJan Kahmen11 min read

Security Testing in the Context of the MITRE ATT&CK Framework

Cyberattacks follow recurring patterns. The MITRE ATT&CK Framework systematically catalogs these patterns and provides organizations with a common language.

Cyberattacks follow recurring patterns. The MITRE ATT&CK Framework systematically catalogs these patterns and provides organizations with a common language to understand threats and prioritize protective measures. Particularly relevant is the "Initial Access" tactic (TA0001), which describes how attackers first gain entry into a network.

This article maps various security testing disciplines to their corresponding MITRE techniques and shows how organizations can systematically assess their attack surface.


External Network Pentest

The External Network Pentest simulates an attacker attempting to penetrate the IT infrastructure from outside. Publicly accessible systems such as firewalls, VPN gateways, mail servers, and exposed services are examined for vulnerabilities.

MITRE Mapping

Technique IDNameDescription
T1190Exploit Public-Facing ApplicationExploitation of vulnerabilities in internet-exposed applications
T1133External Remote ServicesAttacks on VPN, RDP, Citrix, and other remote access services

Scope of Testing

A professional infrastructure pentest includes identification of active services through port scans, analysis of configuration errors, testing for known vulnerabilities such as ProxyLogon or Ripple20, and evaluation of firewall rules and network segmentation. The results serve as a foundation for targeted hardening measures.

Further information: turingpoint - Infrastructure Pentest


Web Application Pentest

Web applications are a preferred target for attackers as they are often directly accessible from the internet and process sensitive data. A Web Application Pentest examines the application layer for security vulnerabilities.

MITRE Mapping

Technique IDNameDescription
T1190Exploit Public-Facing ApplicationExploitation of vulnerabilities in web applications and APIs
T1189Drive-by CompromiseCompromise through manipulated web content

Scope of Testing

The assessment follows recognized standards such as the OWASP Testing Guide and covers authentication and authorization mechanisms, session management, input validation (SQL Injection, XSS, Command Injection), API security, and business logic flaws. Modern web applications frequently use third-party APIs, which should also be included in the assessment.

Further information: turingpoint - Web Application Pentest


Cloud Security Assessment

Cloud infrastructures such as AWS, Azure, and GCP offer tremendous flexibility but also introduce new attack vectors. Misconfigurations in IAM policies, open storage buckets, or insecure service accounts are among the most common vulnerabilities.

MITRE Mapping

Technique IDNameDescription
T1078.004Valid Accounts: Cloud AccountsAbuse of compromised cloud credentials
T1199Trusted RelationshipExploitation of trust relationships (Federated Identity, Service Accounts)
T1190Exploit Public-Facing ApplicationAttacks on cloud APIs and serverless functions

Scope of Testing

A Cloud Security Assessment analyzes IAM permissions for exploitable misconfigurations, storage configurations and data access rights, network security groups and virtual private clouds, logging and monitoring, and compliance with cloud-specific best practices such as CIS Benchmarks. The assessment is conducted within the shared responsibility model, where the customer is responsible for secure configuration.

Further information: turingpoint - Cloud Pentests


Static Code Analysis / SAST

While the previous disciplines uncover vulnerabilities in production systems, Static Application Security Testing (SAST) intervenes earlier: in the source code, before it goes into production. SAST is not an attacker technique in the MITRE sense but rather a preventive measure that complements the framework.

Relation to the MITRE Framework

ReferenceDescription
Prevention for T1190Identification of injection flaws, buffer overflows, and insecure configurations before deployment
Prevention for T1195.001Detection of vulnerable dependencies through Software Composition Analysis (SCA)

Scope of Testing

Static code analysis identifies security vulnerabilities through data flow analysis, taint analysis, and lexical analysis of source code. The assessment covers SQL injections, cross-site scripting, hardcoded credentials, insecure cryptography, and violations of coding standards (MISRA, CERT, OWASP). SAST can be integrated into CI/CD pipelines and provides direct feedback during development.

Further information: turingpoint - Static Code Analysis


Social Engineering Assessment

Technical security measures can be as sophisticated as possible - humans often remain the weakest link. Social Engineering Assessments test how well employees can recognize and defend against manipulative attacks.

MITRE Mapping

Technique IDNameDescription
T1566PhishingAttacks via email with malicious attachments or links
T1566.001Spearphishing AttachmentTargeted attacks with manipulated file attachments
T1566.002Spearphishing LinkTargeted attacks with malicious links
T1566.003Spearphishing via ServiceAttacks via third-party platforms (social media, messenger)
T1566.004Spearphishing VoiceAttacks via telephone (vishing)
T1659Content InjectionManipulation of network traffic to inject malicious content

Scope of Testing

A Social Engineering Assessment includes phishing simulations with realistic scenarios, smishing (SMS-based phishing), vishing (telephone manipulation), pretexting, and measurement of detection rates and response times. More than 75 percent of all social engineering attacks begin with a phishing email. The results serve to raise targeted awareness and build a security awareness culture.

Further information: turingpoint - Social Engineering


Physical Security Assessment

Not all attack vectors are digital. Physical penetration tests examine whether attackers can gain physical access to buildings, server rooms, or endpoints.

MITRE Mapping

Technique IDNameDescription
T1200Hardware AdditionsIntroduction of hardware such as keyloggers, rogue access points, or USB devices
T1091Replication Through Removable MediaSpread of malware via USB sticks and other removable media
T1669Wi-Fi NetworksAttacks via wireless networks, rogue access points

Scope of Testing

Physical security assessments cover access controls (tailgating, badge cloning), server room security, USB drop attacks, WLAN security (evil twin, deauthentication), and verification of clean desk policies. Successful physical access often enables further digital attacks.

Further information: turingpoint - Security Assessments


Mobile Application Pentest

Mobile applications process sensitive data directly on the device and communicate with backend systems. Vulnerabilities can lead to compromise of user data and corporate resources.

MITRE Mapping (Mobile Matrix)

The Mobile Matrix of the MITRE ATT&CK Framework catalogs specific attack techniques for iOS and Android, including malicious apps through sideloading or manipulated app stores, drive-by compromise via mobile browsers, exploitation of MDM vulnerabilities, and insecure data storage and communication.

Scope of Testing

Mobile app pentests follow the OWASP Mobile Security Testing Guide (MSTG) and the Mobile Application Security Verification Standard (MASVS). The assessment covers local data storage and encryption, authentication and session management, network communication and certificate pinning, reverse engineering and tampering protection, and inter-process communication and API security.

Further information: turingpoint - Mobile Application Pentest


Red Team Engagement

A Red Team Engagement is the most comprehensive form of security testing. Unlike traditional pentests, which aim to uncover as many vulnerabilities as possible, the Red Team pursues a defined objective using all available means.

MITRE Mapping

Technique IDNameDescription
T1195Supply Chain CompromiseManipulation of the supply chain (software, hardware)
T1195.001Compromise Software DependenciesManipulation of software dependencies and development tools
T1195.002Compromise Software Supply ChainManipulation of application software before delivery
T1199Trusted RelationshipExploitation of trust relationships with partners and service providers
T1078Valid AccountsUse of compromised credentials

A Red Team Engagement combines all previously mentioned disciplines and emulates the behavior of real attackers, including Advanced Persistent Threats (APT).

Scope of Testing

Red Teaming tests an organization's detection and response capabilities. The team attempts to remain undetected as long as possible while gaining access to sensitive information. The methods employed include technical exploitation, social engineering (physical and electronic), and chaining of multiple vulnerabilities. Red Team Assessments are suitable for organizations with mature security programs that already conduct regular pentests.

Further information: turingpoint - Red Teaming


Conclusion

The MITRE ATT&CK Framework provides a structured foundation for mapping security testing measures to real attack techniques. The combination of different testing disciplines enables a holistic assessment of the security posture:

DisciplineFocusRecommended Frequency
External Network PentestPerimeter securityAnnually or after major changes
Web Application PentestApplication securityWith every major release
Cloud Security AssessmentCloud configurationQuarterly
SASTPreventive code securityContinuously in CI/CD
Social EngineeringEmployee awarenessSemi-annually
Physical SecurityBuildings and hardwareAnnually
Mobile App PentestMobile applicationsWith every major release
Red Team EngagementEntire organizationEvery 1-2 years

The choice of appropriate measures depends on the maturity level of the IT security organization, regulatory requirements, and the individual threat landscape. A structured approach aligned with the MITRE ATT&CK Framework enables prioritization of measures and provides evidence to regulatory authorities and auditors.