Online Influence Campaigns: Strategies and Vulnerabilities

TL;DR

The paper examines inauthentic, societal-scale manipulation as a multi-faceted threat to democratic cohesion, integrating content types, deployment methods, and vulnerabilities across structural, social, and individual levels. It synthesizes case studies (2016, Doppelganger, 2024) to illustrate actor capabilities and AI-enabled tactics, and analyzes the evolving risk posed by AI, including hypothetical AGI/ASI scenarios. The authors advocate for combined technical and policy defenses—content verification, provenance tracking, and governance frameworks—to counter rapid, coordinated manipulation. Their findings stress that safeguarding democratic discourse requires proactive international collaboration and regulation that keeps pace with AI evolution. The work highlights the urgency of mitigating both current and emergent manipulation vectors as digital space becomes more instrumental in shaping public opinion.

Abstract

In order to combat the creation and spread of harmful content online, this paper defines and contextualizes the concept of inauthentic, societal-scale manipulation by malicious actors. We review the literature on societally harmful content and how it proliferates to analyze the manipulation strategies used by such actors and the vulnerabilities they target. We also provide an overview of three case studies of extensive manipulation campaigns to emphasize the severity of the problem. We then address the role that Artificial Intelligence plays in the development and dissemination of harmful content, and how its evolution presents new threats to societal cohesion for countries across the globe. Our survey aims to increase our understanding of not just particular aspects of these threats, but also the strategies underlying their deployment, so we can effectively prepare for the evolving cybersecurity landscape.