← Learning paths

Below the App Layer: Model & Inference Attacks

Technical experts & AI security researchers5 steps · ~41 min

The deepest tier: transformer internals, weight tampering, and inference-time manipulation — the attacks that bypass every application control.

By the end: You can reason about attacks below the application layer — weights, caches, decoding — and explain why app-level guardrails don't see them.

0/3 scenarios completed
Start the path →

AI RiskAtlas is an educational model of how GenAI & agentic systems work and fail. Architectures and payloads are illustrative and simplified for learning — not operational guidance. Real-world cases are summarised from public reporting.

Sources & further reading →·Built by Shi Yuan ↗