•  
  •  
 

Abstract

This Article empirically investigates the connection between harmful narratives on social media and violent acts perpetrated against persons protected under international humanitarian law (IHL) during armed conflict, and then examines the legal implications of such a relationship. First, examining comprehensive datasets from two inter-state armed conflicts (Russia-Ukraine and Armenia-Azerbaijan) as case studies, this Article documents patterns in Telegram data indicating that more violent acts are preceded by increases in harmful narratives than decreases in such content. This suggests that harmful narratives on social media platforms may plausibly contribute to the incidence of real-world violence committed against protected persons during armed conflict. Assuming such a relationship exists, this Article then identifies the legal implications under three relevant bodies of law, arguing that creators and disseminators of such narratives may be in violation of international or domestic law because such content can foreseeably lead to violence. The Article concludes with several policy recommendations for stakeholders, including belligerents, regulating states, social media companies, and humanitarian actors, mapping out a series of actionable steps that could reduce harm against protected persons in conflict settings. Scholars and policymakers must do more to minimize the potential for harm that social media content presents in armed conflict settings.

Annex 1_Detailed Methodology x2.pdf (230 kB)
Annex 1 - Detailed Methodology

Annex 2. Robustness Checks.pdf (132 kB)
Annex 2 - Robustness Checks

Annex 3- Keywords for NK Case.pdf (4119 kB)
Annex 3 - Keywords for NK Case

Annex 4.pdf (177 kB)
Annex 4 - Omitted Harmful Narratives

Share

COinS