Defeating evasive malware with Peekaboo: Extracting authentic malware behavior with dynamic binary instrumentation

Matthew Gaber

doi:10.1016/j.jisa.2025.104290

Defeating evasive malware with Peekaboo: Extracting authentic malware behavior with dynamic binary instrumentation

2025-10-31

0

OA

AI

DOI：10.1016/j.jisa.2025.104290

Original

Save

View PDF

Abstract

En 中文

• Derived 97 anti-analysis techniques used by Windows malware. • Developed Peekaboo, a DBI tool to capture real malware behavior. • Analyzed 18,527 malware and 1,973 benign samples using Peekaboo. • Released a labeled dataset, scripts, and sample hashes for open science. • First to use DBI to defeat evasions and capture API calls, ASM, and network traffic.

Keywords:

Dynamic binary instrumentation

DBI

Malware analysis

Feature extraction

Evasive malware

Sophisticated malware

Defeating evasive malware with Peekaboo: Extracting authentic malware behavior with dynamic binary instrumentation

Abstract

AI Summary

Journal

Organization