AI agents: A privacy- and security-focused view of agentic workflows

Format: Keynote · Required preknowledge: Intermediate · Room: EG01 (WS)
Time: 10:00–10:45
Topics: Data Protection & Privacy, Cybersecurity, AI Governance

What exactly are AI agents? In this talk, we’ll walk through how (many) agentic frameworks actually work. We’ll debunk marketing myths and point out privacy and security risk as well as likely approaches you already know and understand on mitigating that risk. We’ll also dive into some new ideas for better privacy engineering around agents from research and practice.

Speakers: Katharine Jarmul