Your AI Agents Are Overprivileged: The Case for Fine-Grained Authorization
Talk (60 min)
Traditional access control can’t handle contextual, document-level permissions at scale, leaving enterprises vulnerable.
In this talk, discover how Fine-Grained Authorization (FGA) locks down RAG and agentic AI systems, ensuring AI respects strict data boundaries. We’ll dive into real-world implementations with OpenFGA & LangChain, tackling multi-tenant isolation, leakage prevention, and auditability, all while scaling to billions of access decisions without sacrificing performance.
Whether you're building internal copilots or customer-facing AI, this is a must-win security battle.
Ashish Jha
Ashish is a creative technologist and Solutions Engineer at DevRev, where he drives revenue growth and fosters customer partnerships by tackling intricate challenges that lie at the intersection of Customer Experience and Conversational AI-powered developer interactions. Previously at Okta, he deepened his expertise in identity and access management, and he is a leading community voice in Conversational AI, having built over 800 voice applications, won multiple Amazon Alexa Hackathons, and created in-depth training courses. Ashish brings a unique blend of technical depth and user-centric, secure design to everything he builds.