Digital Extremism and Counter-Radicalisation: A Theoretical Account of Algorithmic Pathways, Online Mobilisation and the Limits of Content Moderation in African Contexts

Abstract

Violent extremist movements in Africa recruit, coordinate and propagandise through the same platforms and messaging applications that carry ordinary social life, and the algorithms that sort content for engagement are widely suspected of drawing individuals toward them. This theoretical paper develops an account of digital extremism and its counters in African contexts. Drawing on radicalisation theory and the terrorism-studies literature, on recent audit and experimental research on recommendation algorithms, on the scholarship on deplatforming and platform migration, and on research on content moderation in the Global South and in low-resource languages, it distinguishes three functions digital platforms perform for extremist movements, exposure (bringing content to the susceptible), community (connecting them to one another) and operation (coordinating action), and argues that recommendation algorithms contribute most to the first, moderately to the second and little to the third, which has migrated to encrypted and fringe channels beyond moderation. It shows that content moderation designed in and for the Global North fails in African contexts for reasons of language, context, platform investment and the political ambiguity of what counts as extremism, and that the movements most lethal on the continent mobilise primarily through territorial governance and offline networks rather than through feeds. The paper derives seven propositions about where online intervention can reduce mobilisation and where it cannot, and about the risk that counter-extremism measures expand surveillance and suppress dissent. Nigeria's experience with Boko Haram and the Islamic State West Africa Province anchors the analysis. The paper closes with implications for governments, platforms, civil society and research.