What you'll learn

To secure generative AI, there’s lots that the same as securing any system, but there are also significant differences. In this course, GenAI Security Foundations , you'll learn the basics of generative AI security and discover how you existing skills are still essential. First, you’ll see the architectural components of a GenAI system including the context window, system prompts, and RAG pipelines – these are essential to understand the attack surface. Next, you’ll investigate the three-tier trust hierarchy and how that governs model behavior and determines attack resistance at each layer. Then you’ll discover how to take advantage of the attack surface, with poisoning and with prompt injection. Then, you'll explore the non-deterministic nature of Gen AI and how that affects our traditional security knowledge. Finally, you’ll learn practical takeaways for the common organizational security functions. When you’ve finished this course, you’ll have a good understanding of what’s different with GenAI systems and how you can secure them.