Lightning Talk: AI Penetration Testing
summary
Covers two red team approaches: ethical hacking of others models and adversarial red teaming. Demonstrates file fuzzing via bit-flipping file fuzzing against Whisper which calls ffmpeg, causing a memory exceeded / signal and an ffmpeg segmentation fault traced to a null pointer (not clearly exploitable but proof of concept). Shows creating legal WAVs with added Echo Distortion that breaks transcription and skips or mangles words, implying echo cancellation and input sanitization are required.
AI-written summary · about
chapters (4)
related talks
34mLLM Security 101 - An Introduction to AI Red Teaming
05mLightning Talk: AI Cyber Challenge
53mHow to Break AI Systems (Before Someone Else Does)
37mKeynote - Hacking AI How to Survive the AI Uprising
30mThe Simple, Yet Lethal, Anatomy of an AI Attack
42mHow to attack AI systems (and how to defend them) !!!!
01h 00mHacking and Securing AI Systems: Playing With Fire and Controlling the Flare of AI
33mSecuring Agentic AI With PyTorch: Threat Modeling & LLM Red Teaming in Practice