Foundations of AI Security: classify, map, identify, control and document AI systems, risks, controls and evidence.

Welcome to the AI Security Lab

A practical, hands-on environment for auditors and students learning to review AI systems — classifying architectures, mapping data flows, identifying risks, validating controls, and collecting evidence. Pick a module below to get started.

Lab modules

Guided exercise

Architecture Evidence Lab

Drag-and-drop labs: classify the AI architecture, place components and data zones, map risks across the lifecycle, and build an audit evidence trail. Open lab → Threat modeling

ThreatAtlas

Model AI systems, enumerate threats with OWASP LLM & ML frameworks, and track controls and risk treatment. Open ThreatAtlas → Capture the Flag

RAG HR Assistant (CTF)

A deliberately vulnerable retrieval-augmented HR assistant. Practice attacks such as prompt injection, role bypass and document poisoning. Launch CTF →