Bsides Cymru 2024

Automating Binary Analysis With Machine Learning… and a bunch of scripts
04-27, 12:15–12:45 (Europe/London), Sophia Room - Track 2

Reverse engineering, vulnerability research, binary analysis - all of these approaches and disciplines require skill and take time. This talk dives into supporting the latter, by covering what we can do to automate and accelerate approaches to binary analysis and in getting results, identifying findings, and spotting bugs and vulnerabilities quicker.


Reverse engineering, vulnerability research, binary analysis - all of these approaches and disciplines require skill and take time. This talk dives into supporting the latter, by covering what we can do to automate and accelerate approaches to binary analysis and in getting results, identifying findings, and spotting bugs and vulnerabilities quicker.

During this talk we’ll cover a collection of approaches for accelerating binary analysis, covering a rage of areas from onboarding new binaries, diffing code, and identifying vulnerabilities/ similar code using ML. This will include:

  • Quick wins you can implement right now to accelerate your manual analysis
  • Approaches to developing your own automated approaches to binary analysis
  • Where machine learning fits into this, and a collection of ML automation tooling

James is a software engineer and security researcher, with a background of over six years in the computer security industry - with a primary focus in vulnerability research.