Your AI Agents Are Overprivileged
Your AI helpers might have way more data access than they actually need, which is a huge security no-no. Traditional access controls can’t keep up when you need fine-grained, document-level permissions at scale, leaving your sensitive info vulnerable.
Enter Fine-Grained Authorization (FGA). In his NDC Copenhagen talk, Ashish Jha shows how to lock down RAG and agentic AI systems using tools like OpenFGA and LangChain. You’ll learn how to isolate tenants, stop data leaks, and audit every request—all while handling billions of decisions without slowing down. Whether you’re building internal copilots or customer-facing bots, FGA is your next must-have security move.
Watch on YouTube
Top comments (0)