AI_ACTArt. 72voice-validated
AI_ACT Art72: Art. 72
AI_ACT
AL
Founder at SQUR · last verified 2026-10-06
Regulation text
Providers shall establish and document a post-market monitoring system. The post-market monitoring system shall actively and systematically collect, document and analyse relevant data which may be provided by deployers or which may be collected through other sources on the performance of high-risk AI systems throughout their lifetime, allowing the provider to evaluate the continuous compliance of high-risk AI systems with the cybersecurity requirements.
ATT&CK techniques this article tests · 0
| Technique | Why it maps | Confidence |
|---|
Defending mitigations · 7
| Mitigation | What it does | Confidence |
|---|---|---|
| M1047 | 1. Post-market monitoring requires robust auditing to collect and analyze data on system performance and security events, as mandated by Art. 72. | 100% |
| M1050 | 1. Continuous compliance evaluation necessitates regular vulnerability scanning to identify and address weaknesses in high-risk AI systems throughout their lifetime. | 90% |
| M1051 | 1. Secure configuration management is fundamental for maintaining the cybersecurity posture of AI systems throughout their lifecycle, supporting continuous compliance. | 90% |
| M1031 | 1. Segmenting networks containing AI systems limits the impact of potential breaches and restricts lateral movement by attackers, enhancing overall security. | 80% |
| M1056 | 1. Protecting privileged accounts is critical for preventing unauthorized access and control over high-risk AI systems, a key aspect of cybersecurity requirements. | 80% |
| M1048 | 1. Filtering network traffic detects and blocks malicious communication, supporting the continuous monitoring of AI system security and data integrity. | 80% |
| M1035 | 1. Restricting network access to AI system resources reduces the attack surface and enhances overall security, contributing to continuous compliance. | 80% |
Underlying weaknesses · 7
| CWE | Why it persists | Confidence |
|---|---|---|
| CWE-20 | 1. Lack of input validation can lead to various attacks against AI systems, compromising their performance and security, which post-market monitoring must detect. | 90% |
| CWE-287 | 1. Weak or missing authentication allows unauthorized access to AI systems, directly violating cybersecurity requirements and necessitating continuous verification. | 90% |
| CWE-276 | 1. Default insecure permissions can grant attackers unauthorized access or privilege escalation within the AI system's environment, requiring constant vigilance. | 80% |
| CWE-306 | 1. Unprotected critical AI functions can be abused by attackers, leading to system compromise or data manipulation, which monitoring systems must identify. | 80% |
| CWE-502 | 1. Vulnerabilities in deserialization can lead to remote code execution within AI systems, posing a severe security risk that post-market monitoring should address. | 80% |
| CWE-79 | 1. XSS vulnerabilities in AI system interfaces can allow attackers to inject malicious scripts, impacting user interaction and data integrity, requiring continuous assessment. | 70% |
| CWE-89 | 1. SQL injection can compromise databases backing AI systems, leading to data breaches or manipulation, which post-market monitoring must detect and prevent. | 70% |
What SQUR Covers
Web application + API pentesting for OWASP Top 10, business logic flaws, authentication bypass, injection attacks, and other application-layer vulnerabilities. €1,995 per scan, 24-hour turnaround, EU-only data.
What SQUR Does Not Cover
Internal network pentesting, endpoint security testing, physical security assessments, social engineering, or ICT third-party concentration risk reviews. Engage a complementary provider for those scope items.
Provenance
Mapped Q2.2026 using gemini-2.5-flash · €0.0174 compute · voice-rubric self-validated