Threat Modeling Meets Model Training - Web App Security Skills for AI

Learn how to apply existing web application security knowledge to secure AI and LLM-based applications in this 29-minute conference talk from BSidesSF 2025. Discover how traditional web app security principles translate directly to AI systems, while exploring the unique security challenges that emerge specifically in machine learning environments. Examine common web application vulnerabilities that remain relevant in LLM-based applications and identify the distinct security considerations that are exclusive to AI implementations. Gain practical insights into threat modeling approaches for AI systems and understand how your current security expertise can be leveraged to protect modern AI-powered applications.