Controls

Assess your AI controls

Structured assessments mapped to the CoSAI SRF layers, with crosswalks to CSA AICM, OWASP LLM Top 10, and MITRE ATLAS adversarial technique IDs. Export a gap analysis as PDF.

Controls Assessment

Work through controls mapped to each SRF layer using the CSA AI Controls Matrix (AICM v1.0.3) as the authority — 243 controls across 18 domains. Each control shows its AICM ID, lifecycle phase, and linked MITRE ATLAS technique IDs where applicable. Mark Implemented / Partial / Not done / N/A and export a complete gap report.

CSA AICM v1.0.3 MITRE ATLAS PDF export
Start assessment →
AI Security Controls

Layer-by-layer security controls assessment using the SRF 5-layer model. Includes the OWASP LLM Top 10 with adversarial threat mapping to MITRE ATLAS techniques (AML.Txxxx) — each technique links directly to the ATLAS catalogue. Export a security posture report.

OWASP LLM Top 10 MITRE ATLAS PDF export
Start assessment →

Taxonomy crosswalks included

Controls
CSA AICM v1.0.3

243 controls across 18 domains with EU AI Act, ISO 42001, NIST AI 600-1, and BSI AI C4 crosswalks.

Adversarial threats
MITRE ATLAS

Adversarial ML technique IDs (AML.Txxxx) linked directly to atlas.mitre.org for each relevant control.

LLM risks
OWASP LLM Top 10

The ten most critical LLM application risks mapped to SRF accountability layers and ATLAS techniques.