Hello SAAG, OAuth, WIMSE, SCIM This is a friendly reminder of the Thursday morning 11:15-12:15 AI Agent Security Side Meeting @Hunan room.
We want this side meeting to be real problem-driven, clearing out people's confusions, so the agenda items have included several useful analysis you may find interesting: 1. Evidence-driven AI Agent security attack analysis (such as ClawJacked CVE) 2. Real Multi-Agent-Systems business use-case, services flows, and their proposed security requirement. 3. Some real security challenges from close-up investigation in production systems that actually deployed AI Agents. 4. Attempted definition to AI Agents and summarization of general security gaps. 5. Suggested AI Agent security framework that could point to a lot of useful future works. Contributed by these insightful presentations: * Agentic AI and AI Infra security scenario, risk analysis, existing work and gaps, Presenter Chunchi Peter Liu * A New Model for Authentication and Authorization of AI agent Interactions, Presenter Brian Campbell * AI Integrity and Accountability, Presenter Henk Birkholz * Agent Operation Authorization, Framework and Practices, Presenter Dapeng Liu * Authentication and Authorization Framework for AI Agents, Presenter Daniel King https://github.com/liuchunchi/IETF125-AI-Agent-Security-Side-Meeting We sincerely welcome you to join and discuss, identifying the real gaps to fill. Best regards, Peter and proponents
_______________________________________________ OAuth mailing list -- [email protected] To unsubscribe send an email to [email protected]
