Algorithmic violence: an exploration of the YouTube Recommender Algorithm

Abstract

This article seeks to highlight the complicity of YouTube's recommender algorithm in promoting structural violence. It analyses the successors of the ElsaGate phenomenon, to identify the role of the algorithm in proliferating this violence, and embellishes the notion of algorithmic violence as a means to analyse this phenomenon. It develops a mixed methods framework guided by the principles of analytic autoethnography to draw correlations between the theoretical and practical functioning of the recommender algorithm, through the illustration of a case study of Minecraft Monster School. Aided by textual analysis and autoethnographic methods, algorithmic optimisation is identified as an agent of structural violence in reinforcing inequalities and hierarchies on the platform. It also pins accountability to algorithmic optimisation for being a vehicle of violence, for reinforcing violent content. The analysis suggests the role of the political economy of the platform in promoting sensational and divisive trends, and identifies the mechanisms through which it does so.