This session investigates open-source personal AI agents with a focus on security, looking at how tools like Claude, Open Claw and Hermes are changing the way we work and the risks they introduce. We will walk through a live demonstration of these personal agents in action, highlighting their autonomous features alongside the security and privacy trade-offs that come with granting an AI direct access to your personal data, web browser, and local machine.

We will discuss how security is natively configured within these personal frameworks and examine what potential controls can be layered in to help reduce risk.

The presentation will also create space for group discussion and shared learning around how teams are thinking about agentic workflow security more broadly. Attendees are encouraged to exchange ideas, challenges, and emerging practices for utilizing autonomous personal agents effectively while managing risk.

Kevin has worked in information security in FinTech for the last 7 years, focusing in Application Security, Red Teaming and more recently AI Security. His current focus areas include secure design patterns for AI systems, AI security testing and AI for offensive security.

Registration: RSVP encouraged
Cost: Free
Format: Hybrid