WikipediaBot: Automated Adversarial Manipulation of Wikipedia Articles

TL;DR

WikipediaBot is an automated adversarial system that creates contextually-relevant, hard-to-detect edits on Wikipedia, posing significant threats to article integrity and requiring new detection and defense strategies.

Contribution

The paper introduces WikipediaBot, a self-contained system for generating adversarial Wikipedia edits using linguistic manipulation and credential bypass techniques.

Findings

WikipediaBot can produce contextually relevant adversarial edits.

It can bypass standard login protections and detection mechanisms.

The paper discusses potential defenses against such automated attacks.

Abstract

This paper presents an automated adversarial mechanism called WikipediaBot. WikipediaBot allows an adversary to create and control a bot infrastructure for the purpose of adversarial edits of Wikipedia articles. The WikipediaBot is a self-contained mechanism with modules for generating credentials for Wikipedia editors, bypassing login protections, and a production of contextually-relevant adversarial edits for target Wikipedia articles that evade conventional detection. The contextually-relevant adversarial edits are generated using an adversarial Markov chain that incorporates a linguistic manipulation attack known as MIM or malware-induced misperceptions. Because the nefarious use of WikipediaBot could result in harmful damages to the integrity of wide range of Wikipedia articles, we provide an elaborate discussion about the implications, detection, and defenses Wikipedia could employ to address the threat of automated adversarial manipulations and acts of Wikipedia vandalism.