AI and Africa
Creating AI-driven solutions that meet the real needs of African communities and contribute to sustainable development and improved quality of life. How can people in Africa create AIs by and for their own communities? What makes an AI culturally relevant, resource aware, and context specific? How do we tailor LLMs to specific African contexts and languages?

We are researching AI technologies to tackle critical challenges across Africa. Regions in the Global South, such as those in Africa, are often underrepresented in machine learning literature despite having unique needs and contexts that differ systematically from those in more developed countries. This gap in research can lead to solutions that are either ineffective or unsuitable for the distinct challenges these communities face.
The challenges in the Global South often require solutions that differ from those used elsewhere due to differences in local contexts, resources, infrastructure, and cultural factors. To effectively address these challenges, there is a need for tailored methods that consider the specific conditions and needs of the region. Developing AI solutions that are culturally relevant, resource-aware, and context-specific is crucial to ensuring their success and sustainability in these settings.
One of these initiatives focuses on enhancing the impact of fact-checking organizations in Africa by developing contextualized and locally aware classifiers. By understanding their current strategies and tools for creating and disseminating fact-checked content and examining user responses to these efforts, we aim to develop AI-driven tools that make fact-checkers more effective and efficient.
In another project, we are developing systems that improve a language model’s ability to provide culturally appropriate health information within African contexts. We are creating datasets to evaluate these language models and methods to enhance their ability to address the unique healthcare challenges in diverse African contexts.