Your AI Agents Are Overprivileged: The Case for Fine-Grained Authorization

Generative AI is everywhere, but your bots often end up with carte blanche access to data they shouldn’t see. Traditional access controls can’t keep up with document-level permissions at scale, leaving RAG and agentic AI systems wide open.

In this talk, Ashish Jha shows how Fine-Grained Authorization (FGA) using OpenFGA and LangChain locks down multi-tenant isolation, prevents data leaks, ensures auditability, and handles billions of access checks—without turning your app into molasses. If you’re building any kind of AI copilot or customer-facing agent, this is the security upgrade you need.

Watch on YouTube