Beyond the Chatbot: Why Your AI Agents are Your Newest (and Most Vulnerable) Colleagues

This webinar summary explains how AI agents function like employees and introduce new security risks — including prompt injection, context manipulation, indirect prompt injection, and data exfiltration via crafted content — and describes "Shadow AI" risks. It highlights that many production AI deployments are vulnerable, reviews social-engineering-style attacks against agents, and recommends governance measures (sanctioned/guided/prohibited tools), inventories, least-privilege by task, prompt hygiene training, and human-in-the-loop controls.