Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
166 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Fakeium: A Dynamic Execution Environment for JavaScript Program Analysis (2410.20862v1)

Published 28 Oct 2024 in cs.CR and cs.SE

Abstract: The JavaScript programming language, which began as a simple scripting language for the Web, has become ubiquitous, spanning desktop, mobile, and server applications. This increase in usage has made JavaScript an attractive target for nefarious actors, resulting in the proliferation of malicious browser extensions that steal user information and supply chain attacks that target the official Node.js package registry. To combat these threats, researchers have developed specialized tools and frameworks for analyzing the behavior of JavaScript programs to detect malicious patterns. Static analysis tools typically struggle with the highly dynamic nature of the language and fail to process obfuscated sources, while dynamic analysis pipelines take several minutes to run and require more resources per program, making them unfeasible for large-scale analyses. In this paper, we present Fakeium, a novel, open source, and lightweight execution environment designed for efficient, large-scale dynamic analysis of JavaScript programs. Built on top of the popular V8 engine, Fakeium complements traditional static analysis by providing additional API calls and string literals that would otherwise go unnoticed without the need for resource-intensive instrumented browsers or synthetic user input. Besides its negligible execution overhead, our tool is highly customizable and supports hooks for advanced analysis scenarios such as network traffic emulation. Fakeium's flexibility and ability to detect hidden API calls, especially in obfuscated sources, highlights its potential as a valuable tool for security analysts to detect malicious behavior.

Summary

We haven't generated a summary for this paper yet.

X Twitter Logo Streamline Icon: https://streamlinehq.com