FUTURE FORWARD: Fighting fake news with AI — how west Africa’s Dubawa is transforming fact-checking
Tools developed by a think-tank in west Africa are empowering local media and the public to distinguish between true and false information, especially as it relates to news on their own continent.
In west Africa, the Centre for Journalism Innovation and Development (CJID) in Abuja, Nigeria, is leveraging artificial intelligence to revolutionize fact-checking. Led by head of innovation Monsur Hussain, a team of software developers and AI researchers is working on Dubawa AI, a cutting-edge fact-checking tool launched in 2024. The project aims to empower journalists and the public to distinguish between true and false information, particularly in the context of local news and issues.
Hussain emphasizes that the team initially focused on distinguishing factual claims from questions, with an emphasis on claims that have a significant societal impact, such as health-related and political claims. Nigeria, with over 148 million active internet subscriptions, is Africa's largest online market. As generative AI becomes more sophisticated, creating convincing modified content, journalists face mounting pressure to verify content before publication.
Japhet Johnson, a software engineer based in Abuja, highlights the importance of locally developed AI models that are trained using regional information, which often gets overlooked by global models trained primarily for Europe or the United States. Johnson believes that strengthening local data infrastructure is crucial for African AI systems to remain reliable and compliant with national data protection laws.
Dubawa AI offers a user-friendly chatbot and audio tool, accessible through the Dubawa website and WhatsApp. Users can input claims and receive instant verdicts on the accuracy of the information. The chatbot has already verified over 6,000 claims and has been utilized across 32 African countries, with growing adoption in Ghana, Gambia, Liberia, Sierra Leone, and Senegal.
To train the AI model, the team linked Dubawa's database of verified fact-checks with a machine-learning algorithm that identifies trends in previously reviewed allegations. The developers built a minimum viable product, testing it with journalists and fact-checkers, refining the tool through regular feedback to ensure accuracy and usability. Hussain emphasizes that the ultimate goal is not just to increase the number of users but to foster a culture of critical thinking and verification among the public.
Written by urgent.news from Daily Maverick's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
