A Hacker's Guide to Reducing Side-Channel Attack Surfaces Using Deep Learning

Explore a cutting-edge approach to reducing side-channel attack surfaces using deep learning in this 47-minute Black Hat conference talk. Delve into the effectiveness of deep-learning based side-channel attacks and their automated implementation techniques. Learn how to leverage AI explainability to quickly assess vulnerable parts of an implementation. Follow a step-by-step example demonstrating the approach's potential, limitations, and practical applications. Cover topics including side-channel attacks, whitebox attack models, prediction models, explainability, benchmarking, and results analysis. Gain valuable insights from speaker Elie Bursztein on enhancing cybersecurity defenses against sophisticated side-channel threats.

Intro
Overview
What is SideChannel Attack
WhiteBox Attack Model
Prediction
Model
Explainability
Benchmarking
Results
Back on Track
Outro