Twelve percent of Ethiopia’s population lives in the drought-prone lowlands, where poverty rates are notably higher than in other ecological areas. Predicting oncoming droughts is crucial for mitigating their damaging effects on health, livelihoods and ecosystems. Researchers at Addis Ababa University and The University of Gondar will be developing a drought early warning system, using machine learning, a technology that previous research in this area has not fully leveraged. The work will focus on four woredas (districts) in Afar and Somali regions with high vulnerability to drought-induced poverty and will utilise satellite imagery and weather station data to provide accurate drought prediction models. Researchers plan to create a user-friendly Al-enabled dashboard which can be accessed by government agencies and local communities – enabling policymakers to design targeted responses to drought challenges and farmers to implement drought-resistant practices. Findings will also inform the government’s wider efforts to develop effective poverty alleviation and disaster recovery strategies.