Hacking Machine Learning Systems (Red Team Edition) - AI Hacker

This video will get you up to speed on red teaming and hacking of machine learning systems (and how to defend them) in a practical fashion.
The content covers both typical infrastructure attacks and also adversarial machine learning attacks - and in addition practical mitigation strategies are discussed.
This presentation was part of Grayhat 2020 - Red Team Village, named "Learning by doing: Building and breaking a machine learning system".
Enjoy!
Intro 0:00
Getting started resources 3:28
Fake-Signing Malware to bypass AV models (Microsoft Evasion Competition) 4:37
Building Husky AI and the Machine Learning Pipeline 8:32
Neural Networks - Basics 15:36
Operationalizing the machine learning model 20:13
Threat Modeling 22:01
Selected Threats to explore 26:00
Attacks to misclassify images (Bruteforce, Random Pixels, Fast Gradient Sign,...) 29:00
Image Rescaling Attacks 39:00
Stealing the model file 41:07
Backdooring a machine learning model 44:44
Generative Adversarial Networks 50:28
Microsoft VS Code - CVE Details 55:50
Conclusions 57:29

[@octopus3141]

Great stuff 👍

[@embracethered]

Thanks for the visit and note. Appreciate it! Let me know if there are any relevant topics you'd like to see covered?

[@brunao9689]

Great video man! Hope u reach 1000 soon

[@embracethered]

Thank you, really appreciate the note! 🙂

[@Dani1989474]

i love this video. i will keep an eye out for you from now on :)

[@embracethered]

Thank you! Really appreciate the comment!

[@th3pac1fist]

🔥

[@embracethered]

Thanks!! It's probably one of my most interesting videos.

[@Active-AI]

Great stuff, thanks!

[@embracethered]

Glad you liked it. Thanks!