Part 4: Generative AI and Adversarial Attacks
In this part, you will learn the fundamentals of generative AI and how it differs from classic predictive AI. You will learn how to develop Generative Adversarial Networks (GANs) and how you can use them or pre-trained GANs to stage adversarial attacks or create deepfakes. You will learn how to develop chatbot applications with ChatGPT and LangChain and how to stage prompt injection attacks, poison RAG, embeddings used in RAG, and fine-tuning. You will also learn how to stage poisoning attacks on open source models on Hugging Face, model lobotomization, and how privacy attacks apply to LLMs.
This part has the following chapters:
- Chapter 11, Generative AI – New Frontier
- Chapter 12, Weaponizing GANs for Deepfakes and Adversarial Attacks
- Chapter 13, LLM Foundations for Adversarial AI
- Chapter 14, Adversarial Attacks with Prompts
- Chapter 15, Poisoning Attacks and LLMs
- Chapter 16, Advanced Generative...
