🖍️
Pesidious
  • Introduction
  • Pre-requisites
  • Train Models (optional)
  • Mutate Malware
  • HowTo
    • Design your own RL agent
    • Train models against your custom classifier
    • Add more mutations
  • Conclusion
Powered by GitBook
On this page
  • Step 1: Get the original scores of malware samples
  • Step 2: Mutate malware files
  • Step 3: Check the score for mutated malware

Was this helpful?

Mutate Malware

PreviousTrain Models (optional)NextDesign your own RL agent

Last updated 4 years ago

Was this helpful?

The tool can be used to create mutated malware that can evade an AI-powered classifier.

Prerequisites

  • Environment

This tool is designed to work with only 32 bit PE binaries.

Step 1: Get the original scores of malware samples

In order to assess the evasiveness of the mutated malware generated by the tool, you can first run the classifier against the malware samples. 

python classifier.py -d /path/to/directory/with/malware/files

Step 2: Mutate malware files

The tool comes with pre-trained models that can mutate a malware in order to evade the in-built classifier.

Run the following command with to mutate malware files

python mutate.py -d /path/to/directory/with/malware/files

The mutated malware files will be stored in a directory called Mutated_malware in the following format

Mutated_malware/mutated_<name-of-the-file>

Step 3: Check the score for mutated malware

You can again run the classifier but against the mutated malware and compare it with the original scores retrieved in the first step

python classifier.py -d Mutated_malware/
Train Models (optional)