Explore real-world insights and practical lessons from building and defending Large Language Model (LLM) applications in this 27-minute conference talk from DEF CON 32. Gain firsthand knowledge about critical vulnerabilities encountered during development, innovative defense mechanisms implemented, and essential strategies for strengthening AI application security. Learn actionable approaches to navigate the complex landscape of AI security, drawing from direct experience in developing and protecting LLM systems in production environments. Master practical techniques for enhancing the robustness of AI applications while understanding common pitfalls and their solutions in LLM implementation and security.
Lessons Learned from Building and Defending LLM Applications
Overview
Syllabus
DEF CON 32 - Lessons Learned from Building and Defending LLM Applications - Javan Rasokat
Taught by
DEFCONConference
Related Courses
-
Red Teaming LLM Applications
-
Building Production-Ready Apps with Large Language Models
-
Securing AI and LLM-based Applications: Best Practices
-
LLM SecOps - Securing Large Language Model Applications
-
Ethical and Secure AI Vulnerability Disclosure - Challenges and Best Practices
-
NeuroCTI - A Custom Large Language Model for Cyber Threat Intelligence: Benchmarking, Successes, Failures and Lessons Learned
