Modern Appsec vs. GenAI Application: Is Your Appsec Ready?
summary
Maps OWASP LLM Top 10 onto a modern AppSec stack and shows which risks are already handled and which need new work: prompt injection and hallucination/over‑reliance require model tuning and runtime checks, RAG needs secured training data and a RAG/Vector DB review, agents and plugins demand privilege separation, and code generation produces insecure or outdated libraries so 60% deploy without testing and lack of monitoring are real surprises; recommends cod‑gen governance, enforced full reviews, modern SCA and developer training.
AI-written summary · about
chapters (8)
- 0:00Modern AppSec vs GenAI apps
- 6:06Modern AppSec program overview
- 10:00High-level architecture of generative AI applications
- 13:53OWASP LLM Top 10 mapping
- 26:35AI code security risks
- 33:45Mitigating AI code vulnerabilities
- 41:55Real-world breaches and code generation security
- 46:30Q&A
related talks
35mGenAI Application Security: Not Just Prompt Injection
40mState of (Absolute) AppSec
30mNavigating the New AI Security Landscape
49mAppSec, Ready, Set, Go!
59mSecuring the Digital Landscape: A Deep Dive into OWASP Top 10 for Applications, APIs, and LLMs
43mThe State of AppSec in the Era of AI [Panel]
51mShifting Knowledge Left: Keeping up with Modern Application Security
55mUsing GenAI on and inside your code, what could possibly go wrong?