Sudan’s civil war has produced one of the world’s worst humanitarian catastrophes. Since fighting erupted in April 2023 between the Sudanese Armed Forces (SAF), led by Abdel Fattah al-Burhan, and the paramilitary Rapid Support Forces (RSF), led by Mohamed Hamdan Dagalo (Hemedti), tens of thousands have been killed, millions displaced, and warnings of atrocity crimes and genocidal violence have intensified, particularly in Darfur. Yet the war has unfolded alongside another conflict: a vast online struggle over narrative, legitimacy, and visibility.For years, social media companies have claimed they are doing a lot to tackle disinformation, and bot networks (or electronic flies as we call them in the Gulf). Networks are supposedly identified, exposed, and removed. Yet a recent investigation I conducted as a UNESCO Fellow into influence operations surrounding Sudan uncovered something troubling: a sprawling bot and sockpuppet network that has operated across the Middle East for years, surviving suspensions, adapting to platform changes, and shaping political discourse across multiple countries and conflicts.The operation consisted of hundreds of core sockpuppet accounts supported by thousands of auxiliary bots producing more than 170,000 posts across multiple languages over several years. Its activity stretched across Sudan, Yemen, Tunisia, Mauritania, Libya, Syria, Egypt, and the Gulf - including Qatar. Sudan sat at the centreof the network’s activity.The findings suggest these networks increasingly function as parallel media infrastructures. They do not simply spread isolated pieces of false information. They sustain broader political narratives through repetition, selective omission, coordinated amplification, and the simulation of public consensus. Collectively, they manufacture the appearance of organic political sentiment while concealing the centralised and inauthentic nature of the operation itself.The network’s activity closely tracked developments in Sudan’s civil war. As violence escalated following April 2023, Sudan-related content surged dramatically. The operation overwhelmingly promoted narratives aligned with the RSF, portraying the group as humanitarian, peace-oriented, and legitimate, while assigning responsibility for civilian suffering, famine, and instability almost exclusively to the SAF and alleged Islamist actors.This framing extended well beyond ordinary wartime propaganda. The RSF leadership was repeatedly associated with diplomacy, civilian protection, and humanitarianism. Reports of abuses attributed to the RSF were systematically ignored, denied, or reframed. The result was a highly disciplined narrative environment in which one side of the conflict was persistently rehabilitated while responsibility for violence was externalised elsewhere.These dynamics matter because information manipulation directly affects how conflicts are understood internationally. In atrocity contexts, coordinated influence operations can muddy attribution, distort public understanding, weaken the visibility of credible reporting, and interfere with humanitarian and diplomatic responses. Sudan’s war has already suffered from limited media access, fragmented reporting, and competing geopolitical interests. Large-scale influence operations further degrade an already fragile information environment.One of the most striking findings in the report concerned the structure of the network itself. Rather than functioning as a single unified cluster, the operation was organised into regional “cells.” Different groups of accounts posed as users from specific countries, often using local flags, dialects, imagery, and national symbols to enhance credibility. Tunisian-focused accounts interacted heavily with other Tunisian-focused accounts. Sudan-focused accounts clustered around Sudanese themes. Mauritanian-focused accounts amplified Mauritanian content.Longitudinal analysis, however, revealed repeated interaction with the same low-salience “filler” accounts focused on entertainment, viral videos, and generic social media content. These auxiliary accounts acted as connective infrastructure linking otherwise separate regional clusters together. What initially appeared to be dispersed online communities increasingly resembled a coordinated transnational influence architecture.Temporal analysis also exposed patterns inconsistent with authentic geographically distributed behavior. Accounts claiming to represent users from multiple countries such as Libya or Tunisia displayed similar posting rhythms aligned with Gulf-region working hours. The further the claimed location was from Gulf Standard Time, the more pronounced the discrepancy became.Perhaps most importantly, the network adapted over time. Accounts were suspended, repurposed, renamed, and replaced. Narratives shifted in response to regional developments. More recently, the operation appears to have incorporated AI-assisted accounts, synthetic personas, and increasingly automated forms of engagement. Some accounts displayed homogenised language patterns, repetitive slogans, and coordinated interaction structures characteristic of emerging AI-enabled influence operations.This persistence is significant. The dominant assumption surrounding online influence operations is often that they are temporary campaigns tied to specific events or elections. What emerged in this investigation was something more durable and infrastructural. The network operated across years, multiple countries, several languages, and different political contexts while maintaining broad narrative consistency and operational coordination.The broader implications extend well beyond Sudan. For years, discussions around coordinated influence operations have focused heavily on Russia, China, or Western elections. Yet the Middle East is one of the world’s most sophisticated environments for digital authoritarianism, computational propaganda, and coordinated inauthentic behaviour. Governments, proxy actors, political networks, and commercial operators across the region have developed increasingly advanced methods for shaping online discourse and manufacturing legitimacy at scale.At the same time, social media platforms continue to struggle with systematic enforcement, particularly in Arabic and other non-English contexts. Many of these networks persist for years despite public exposure. Verification systems designed to enhance trust can instead be exploited to increase visibility and perceived legitimacy. Indeed, 40 of the accounts in this network had verification status, but were entirely fake.The consequences may extend even further as AI systems increasingly retrieve and synthesize information from social media ecosystems. Coordinated influence operations don’t just shape what users encounter online, but they also shape the informational environment from which AI systems generate knowledge itself.While disinformation is a central issue, perhaps more worrying is the emergence of durable influence infrastructures engineered to manipulate visibility, simulate public consensus, and shape political reality across entire regions over long periods of time.• The writer is an associate Professor of Media Analytics at Northwestern University in Qatar, a Unesco Fellow, and author of several books including Digital Authoritarianism in the Middle East.
admin
