AI-Driven Code Obfuscation: Enhancing Software Security using Machine Learning

Conference papers

Authors

Edra Tabaku , Alexander Iliev , Tugce Balli and Kendrick Bollens

Abstract

Code obfuscation is a crucial technique in software security, making code harder to understand and reverse engineer while maintaining its functionality. This research conducts a comparative and experimental study of traditional rule-based code obfuscation tools that use variable renaming, dead code injection, control flow flattening and other techniques, and machine learning tools such as LLMs, neural networks and transformers to obfuscate JavaScript code. The study aims to evaluate the current state of AI code obfuscation and address its limitations.

Keywords

code obfuscation, artificial intelligence, machine learning, JavaScript