TL;DR
This study investigates how the weaponization of social media during the Northern Ethiopia War (2020–2022) fueled intense ethnic polarization and offline violence, while systematically excluding women from both digital discourse and formal peace processes. By applying social identity, liberal feminist, and intersectionality theories, the authors highlight a critical "dearth of digital peacebuilding" in the Global South and call for data-driven, gender-inclusive technological interventions.
What Problem Does This Paper Address?
While social media platforms hold immense potential for democratic mobilization and community building, they are frequently weaponized during civil strife to spread disinformation, incite ethnic hatred, and deepen societal fractures. Most existing literature on online polarization focuses heavily on the Global North and relies primarily on quantitative metrics.
This paper addresses the lack of qualitative, context-embedded research in sub-Saharan Africa by examining the Northern Ethiopia War as a case study. The authors explore how digital polarization directly translated into offline atrocities and analyze why women—who bore the disproportionate brunt of conflict-driven violence (including war rape, displacement, and economic collapse)—were systematically marginalized from both online public spheres and formal, top-down peace negotiations.
Methods
The authors employ a qualitative exploratory research design to capture nuanced, context-specific insights. Their methodology combines three primary data sources:
- Semi-Structured Interviews: 10 key informant interviews (KIIs) conducted with experts and directors from relevant Ethiopian government ministries (e.g., Ministry of Peace, Ministry of Women and Social Affairs) and non-governmental organizations.
- Focus Group Discussions (FGDs): Two separate focus groups, including sessions with the Information Network Security Administration (INSA) and Positive Peace Ethiopia, ensuring a diverse range of participant ages and genders.
- Document Analysis: Evaluation of annual and semi-annual reports from the Ethiopian Human Rights Commission and social media usage data from the Ethiopian Media Authority.
The data was analyzed using an iterative, non-linear thematic coding procedure (spanning open, axial, and selective coding) to discover new concepts directly from the participants' testimonies rather than forcing data into rigid pre-existing frameworks.
Datasets
As a qualitative study, this paper does not utilize standard machine learning datasets or quantitative benchmarks. Instead, the primary "dataset" consists of qualitative text data generated from the 10 key informant interviews and 2 focus group discussions across various governmental and non-governmental organizations in Ethiopia, supplemented by institutional reports.
Key Insights and Ideas
- The "Us versus Them" Algorithmic Multiplier: Social media platforms (specifically Facebook, X, and YouTube) acted as virtual battlegrounds. Algorithms prioritized sensational, divisive content, which reinforced in-group loyalty and facilitated extreme out-group hostility. This digital dehumanization (e.g., using derogatory terms like "junta" and "day-hyena") directly legitimized and escalated offline violence.
- The Vicious Cycle of Exclusion: A stark gender bias was embedded within the peacebuilding architecture. The peace initiatives (such as the Pretoria and Nairobi agreements) followed a strict, masculine-centered, top-down approach that completely omitted female representation, silencing the very demographic most affected by the war's atrocities.
- The Minority Digital Dictatorship: Despite the conflict's massive offline impact, only 5.5% of the Ethiopian population actually uses social media. A small minority of digitally connected online actors disproportionately shaped narratives that destabilized the peace of the non-connected wider population.
- Hostile Online Environments: Women faced a "double marginalization" in the digital sphere. Beyond the broader digital divide and high rates of digital illiteracy, online spaces were highly hostile, featuring gender-based threats and the weaponization of rape reports, which discouraged women from participating in digital peace advocacy.
Key Results
- Thematic Architecture: The qualitative analysis successfully mapped 94 distinct open codes into 25 axial codes, culminating in 5 core selective themes detailing the intersection of social media weaponization, women's marginalization, and the challenges of digital peacebuilding.
- Zero Female Representation: The study confirms that not a single woman directly participated in or signed the high-profile peace initiatives in Pretoria and Nairobi, illustrating a complete exclusion of women from the formal resolution process.
- The Dearth of Digital Peacebuilding: Key informants noted that the active role of digital media in the actual peacebuilding process was "almost zero." Technology was heavily leveraged for warfare and propaganda, but virtually unused for systematic conflict resolution, automated content moderation, or counterspeech.
Limitations and Future Work
The authors acknowledge several limitations in their current work:
- Lack of Firsthand Perspectives: The study relies on institutional experts and directors rather than capturing direct, firsthand accounts from war victims, internally displaced persons (IDPs), or combatants.
- Subjective Bias: The qualitative nature of the interviews means the findings are susceptible to the individual impressions and potential biases of the participants regarding digital echo chambers.
Future Directions: The authors recommend that future research incorporate large-scale quantitative measures, such as social media network analysis, to map algorithmic bias and echo chambers. Furthermore, they advocate for the deployment of advanced AI technologies—such as Natural Language Processing (NLP) for topic clustering, sentiment analysis, and automated hate-speech detection—to build proactive, gender-sensitive digital peacebuilding frameworks tailored to the Global South.
Summary updated