Portfolio News

Empowering Ethiopian Youth to Fight Misinformation

DABBAL Data Science+AI Ethiopia

Jun 17 , 2026

Data Science+AI

DABBAL

Ethiopia

Amount invested $30,450 USD Funding Status active early period Founded in 2021 by Abdurezak Yisihak, Abdurhman Abrar Abdela, Mohammed Ibrahim, Abduljebar Sani & Ajaib Mohammed Adem

We are not trying to tell young people what to think. We are helping them build the habit of questioning emotionally manipulative content before it spreads. UNICEF’s support allows us to build open tools that help young people recognize misinformation and navigate digital spaces safely.

DABBAL

The Problem We Are Addressing

False information in Ethiopia is creating serious social and public health risks. It spreads rapidly through social media and messaging platforms, reaching millions who often lack the tools to critically evaluate what they see online.

96.3% of surveyed Ethiopian digital media users reported encountering misinformation online (Internews, 2023) — with young people among the most vulnerable.

The consequences are not abstract. One of our team members witnessed this firsthand when his father suffered a stroke after following false claims about a traditional diabetes cure circulating online. At the same time, on university campuses across Ethiopia, misinformation has inflamed protests and created tensions with severe consequences for students.

These are not isolated incidents. Misinformation about elections, health treatments, ethnic groups, and community safety spreads daily through WhatsApp forwards and Facebook posts — often shared by young people who trust what they receive from friends or family. Many do not realize something is false until harm has already occurred: a medication taken, a conflict sparked, a vote cast based on fabricated claims. Many young people in Ethiopia receive news primarily through social platforms and peer-sharing networks rather than verified institutional channels. In fast-moving emotional situations, information is often trusted based on who shared it rather than whether it is verified.

When misinformation goes undetected, the social and emotional costs compound. Young people who unknowingly spread false content face damaged reputations when the truth emerges. Communities fracture along lines drawn by fabricated narratives. And when individuals cannot trust what they read online, democratic participation erodes.

Those who benefit most from solving this problem are young Ethiopians, community health workers, educators, and peacebuilders. Those with least to gain — at least in the short term — are those who deliberately weaponize misinformation for political or economic ends.

Introduction

Dabbal Software Development PLC is an Ethiopian technology company building digital solutions for real-world challenges across Africa. Our latest initiative, Horizon Truth, is a platform designed to help young people identify and resist misinformation online.

Horizon Truth combines gamified digital literacy training, crowdsourced content verification, and AI-powered analysis to help users learn how misinformation spreads — and how to stop it. By simulating real-world scenarios such as viral posts, forwarded messages, and manipulated news stories, the platform trains users to evaluate credibility, verify sources, and report suspicious content.

Our mission is to equip Ethiopian youth with the knowledge and tools they need to navigate digital spaces safely and responsibly.

Our Team

Horizon Truth is developed by a team of Ethiopian engineers, researchers, and innovators working at the intersection of AI, digital literacy, and social impact. Our team includes expertise in:

Artificial Intelligence and Machine Learning
Natural Language Processing for low-resource languages
Full-stack software development
Cloud and scalable system architecture
Digital education and community engagement

Our diversity — across disciplines, backgrounds, and technical specializations — strengthens the platform. It means Horizon Truth reflects the realities of Ethiopian communities, not just the assumptions of its builders.

DABBAL

Why Frontier Technology Matters

Misinformation adapts fast and spreads at scale. Manual verification alone cannot keep up. Horizon Truth uses AI and data-driven analysis to identify patterns of misinformation and assist moderators in reviewing reported content.

The platform's AI system analyzes user-submitted reports, detects potential misinformation patterns, and assigns trust scores to prioritize content for human review. This combination of automation and community participation allows faster response to harmful content — while also teaching users how misinformation works.

Why Local Design Matters

Standard English-language tools miss critical context in the Ethiopian information environment. Misinformation in Amharic and Afaan Oromo often travels through culturally specific metaphors, idiomatic expressions, and shared references that global platforms are not built to detect.

Humor and satire, in particular, can be misclassified by models trained on Western data. A phrase that is clearly ironic in Oromo culture may appear genuine to an AI trained primarily on English content — and vice versa. Local contributors who understand these nuances are essential to building datasets that reflect reality on the ground.

Our datasets are sourced and validated in partnership with local communities, researchers, and institutions like Jimma University. This grounding ensures our AI tools serve Ethiopian users — not just approximate them.

What Makes Horizon Truth Unique

Many initiatives focus only on fact-checking or media literacy education. Horizon Truth integrates multiple approaches into a single platform:

Gamified Learning: Users practice identifying misinformation through interactive scenarios that simulate real digital environments — building critical thinking as a habit, not a chore.
Crowdsourced Reporting: Communities can flag suspicious content and learn from verified cases, turning users into active defenders of their information environment.
AI-Supported Detection: Intelligent tools help moderators prioritize and analyze reported content at scale.
Localized Design: Full support for Amharic and Afaan Oromo, ensuring the platform is genuinely accessible — not just translated.

This combination positions Horizon Truth as both a digital literacy platform and a community-driven early warning system against harmful misinformation.

The Gamification Approach

The game mechanics in Horizon Truth are designed to reinforce specific behaviors: source verification, lateral reading, emotional regulation when encountering alarming content, and the habit of pausing before sharing. These are not passive lessons — users practice them repeatedly in simulated high-stakes situations.

When users make a wrong choice in the game, they are shown exactly why the content they evaluated was misleading — the specific technique used, the red flags they missed, and what a reliable source would look like. This immediate corrective feedback is central to building durable recognition skills, not just engagement.

From Individual Behavior to Peacebuilding

When youth identify misinformation earlier, they do something powerful: they stop the chain of amplification. A false rumor about an ethnic group or a political event that would have been forwarded to fifty contacts instead stops with one informed user.

In polarized environments, misinformation spreads fastest when content triggers fear, outrage, or identity-based reactions. Teaching youth to pause and verify before reacting interrupts that escalation cycle at an early stage. This behavior — pausing, verifying, and choosing not to share — is a direct contribution to reducing polarization and conflict. It is small at the individual level and transformative at scale. For impact to compound, however, other actors must also change: educators must integrate digital literacy into curricula, platforms must reward accuracy over engagement, and community leaders must model critical thinking publicly.

DABBAL

Progress and Achievements

In Q1, our team focused on building the platform's foundation and testing its early learning experience with real users.

Developed and refined interactive scenarios simulating common misinformation situations on social media.
Built and launched a prototype web-based platform.
Conducted pilot playtesting sessions with approximately 50 youth in Jimma.
Began developing the core structure of the crowdsourced reporting system.
Improved the UI based on usability testing and began open-source documentation and privacy compliance processes.

Feedback from our 50 pilot testers was direct and valuable. Users flagged confusing navigation flows, requested stronger visual contrast in the game interface, and — most usefully — told us that when they made wrong choices, the explanations weren't clear enough. They wanted to understand why their answer was wrong, not just that it was. That feedback has driven significant redesign of our post-answer explanation screens.

During one pilot session, a student initially trusted a fabricated social media post because it included emotional language and a fake “breaking news” label. After completing the scenario and reviewing the explanation, the same student later reported questioning a similar post shared in a family WhatsApp group — instead of forwarding it immediately. Moments like this suggest the platform is not only improving gameplay performance, but influencing real-world digital behavior. Some of what users did surprised us in the best way. Several participants spontaneously showed the platform to classmates who hadn't been invited to the pilot. Others came back the following day without any reminder — drawn by the challenge of improving their scores. One student told us she had started questioning WhatsApp forwards she received at home, using the reasoning frameworks from the game. That moment — when the platform's logic transfers into real life — is what we are building toward.

The strongest signal of commitment so far has come from Jimma University, which has expressed interest in integrating Horizon Truth into its student orientation programming. That institutional endorsement reflects confidence in both the platform and its approach.

Crowdsourcing & Trust Safeguards

The crowdsourced reporting system now in development will allow users to flag suspicious content for review. But open reporting systems carry risks — abuse, politically motivated mass reporting, and simple crowd error among them.

Our approach to trust and safety is built around several principles. Reporting access will be tiered: verified users with demonstrated track records will carry more weight than anonymous submissions. AI analysis will cross-check reports for coordinated patterns that might indicate brigading or organized manipulation. No content will be actioned based on crowd reports alone — human moderators, informed by community context, will have final review authority.

Dispute resolution will be transparent: users will be able to see why a report was upheld or dismissed. For contributors operating in sensitive or politically polarized contexts, the platform will offer anonymization options and clearly communicate what data is and is not stored.

What happens if the crowd is wrong? Content does not get removed — it gets flagged for moderator review. The system is designed to surface, not to censor. Getting that balance right is one of the most important governance challenges we are working through, and we welcome continued dialogue with partners and advisors on the frameworks that will guide it. As we move into Q2, our focus includes designing the governance framework for the crowdsourced reporting system. This includes contributor reputation models, moderator escalation pathways, transparent review logs, and safeguards against coordinated manipulation. Our goal is to ensure the platform promotes accountability and trust without enabling censorship or political misuse.

DABBAL

How UNICEF's Venture Fund Investment Will Help

The support from the UNICEF Venture Fund will enable us to accelerate development and scale. Specifically, the investment will help us:

Develop the AI-powered misinformation detection and trust scoring system
Localize the platform fully in Amharic and Afaan Oromo
Launch a mobile version optimized for low-bandwidth environments
Expand testing and onboarding to 10,000 youth users across Ethiopia
Conduct digital literacy outreach in schools and communities

Through partnerships with Jimma University and the Ministry of Peace, we aim to integrate Horizon Truth into youth engagement and peacebuilding initiatives at scale.