Fine-Grained Authorization: The Missing Piece in Agentic AI Security
summary
Shows how to enforce document‑level permissions for RAG and agent workflows using OpenFGA plus LangChain and vector DBs (demo uses Chroma). Compares post‑filtering vs pre‑filtering and argues pre‑filtering prevents early leakage, describes task‑level agent permissions so agents start with zero rights and request scoped approval, and surfaces audit logs and multi‑tenant isolation patterns claimed to scale to billions of decisions (play.fga.dev and code examples included).
AI-written summary · about
chapters (11)
- 0:00Introduction
- 3:19Agent security boundaries
- 8:20Slow AI adoption due security
- 13:00Sensitive info disclosure risk
- 17:12Authorization models evolution
- 26:27ReBAC for AI apps
- 31:00OpenFGA implementation
- 36:18RAG assistant permissions
- 43:01RAG security filtering
- 49:01OpenFGA for agent use cases
- 53:36Summary
related talks
30mFine-Grained Authorization: The Missing Piece in Agentic AI Security
30mYour AI Agents Are Overprivileged: The Case for Fine-Grained Authorization
55mYour AI Agents Are Overprivileged: The Case for Fine-Grained Authorization
07mOpenFGA For Agents: Safe Delegation In 5 Minutes
26mBeyond the Gateway: Fine-Grained Authorization for AI Agents
23mZero-Trust Architecture for Agentic AI: Securing Multi-User Access and Third-Party Integrations
54mFine-Grained Authorization for Modern Applications
33mHow to Prevent AI Agents from Accessing Unauthorized Data