About this talk
This talk, presented by Kinnaird McQuade, explores the vulnerabilities associated with AI agents and chatbots that utilize Python code interpreters, particularly in the realm of data analysis. The speaker explains how attackers can exploit these systems through prompt-to-RCE, credentials abuse, and command and control payloads. Additionally, the session covers a case study on an AWS AgentCore breakout and offers strategies for enhancing security to defend against code interpreter exploits.
More from this event
See all 91 talks →
BSidesSF 2026 - Opening Remarks (Sunday) (Reed Loden)
14:36
BSidesSF 2026 - Follow the data to learn the secret (Dylan Ayrey)
35:17
BSidesSF 2026 - Not My Vibe: When AI Coding Agents Go Off the Rails (Aonan Guan, Zhengyu Liu)
45:56
BSidesSF 2026 - "Ask the EFF" Panel (Panel)
45:30