What IDOR (Insecure Direct Object Reference) is, how attackers reach other users' data by swapping an id, what it leads to and how to defend. A practical breakdown.
What agentic AI threat modeling is and how to map trust boundaries, tools and actions for autonomous agents before you build — using OWASP Agentic and the MAESTRO approach.
What an LLM jailbreak is, the main techniques (role-play, encoding, multi-turn), and how to defend the model. Why guardrails are a barrier, not a wall.
How RAG attacks work and how to protect your AI knowledge base: document isolation between users, filtering uploaded content against indirect injection, securing the vector store.
Seven practical rules to secure AI agents with access to tools and APIs: least privilege, tool isolation, action confirmation, defense against indirect prompt injection and monitoring.
What prompt injection is, how direct and indirect attacks work, what they lead to, and how to defend your LLM application. The number-one risk in the OWASP LLM Top 10.
A practical breakdown of the OWASP Top 10 for LLM Applications 2025 — prompt injection, sensitive data disclosure, excessive agency and more — with how each is attacked and defended.