Scott Bell

As GenAI moves from experiment to production, security cannot be an afterthought. This session empowers data professionals to lead the charge in securing LLMs without needing to be cybersecurity experts. Learn how to facilitate a practical Threat Modeling Workshop, a collaborative exercise to identify risks like prompt injection and data leakage,and seamlessly integrate these security checks into an AIOps strategy.

As the industry pivots from Chatbots to Agents, the threat landscape shifts from misinformation to unauthorized action. Danger in Delegation explores the dark side of autonomous AI, dissecting the "Lethal Trifecta" of agentic risks, UI redressing (clickjacking), and the catastrophic consequences of granting LLMs the power to traverse the web and execute system commands on our behalf.