AI introduces powerful capabilities but also novel risks, vulnerabilities, and threats. In this session, Rhys Arkins, VP Product at Mend.io, explores emerging security issues in AI infrastructures, including malicious models, prompt injection, RAG poisoning, and system prompt leakage. Through practical examples and real-life scenarios, attendees will learn how to detect zero-day vulnerabilities, remediate threats proactively, and effectively implement holistic safeguards. Equip yourself with actionable strategies to manage AI risks before they flare out of control.