AI for climate-health crises: analysis from Ethiopia's Somali region

Abstract

Climate related shocks cause severe risks to health systems in Ethiopia’s Somali Region, particularly among vulnerable and displaced pastoralist communities. This note discusses how an artificial intelligence (AI) driven evidence synthesis system, co-designed with local stakeholders, could strengthen responses to climate-related health crises in this fragile setting. Stakeholder consultations and participatory co-design workshops were conducted for the development of an AI-enabled evidence synthesis platform (Solace-AI). Participants included vulnerable and displaced communities, regional and national health officials, disaster risk management authorities, and international research partners. Climate catastrophes compounded by decades of conflict-related trauma, mental health problems have become one of the most pressing health issues in the region. Understanding and addressing this interconnected crisis, in which the effects of environmental catastrophe, forced migration, mental illness, and substance misuse are compounded, it is essential for any meaningful intervention in the region. Decision-makers, from national ministries to regional disaster management offices to local health workers, need timely access to evidence to inform their actions. In settings already strained by climate impacts, commissioning formal evidence syntheses is often impossible. AI-enabled evidence synthesis could improve preparedness and responses, thus decision-makers must be capacitated to proceed with whatever information is immediately available, which may be incomplete or not contextualized to their setting. However, effectiveness depends on participatory co-design, ethical governance, sustainability, and local ownership. Without these safeguards, AI risks reinforcing existing inequities rather than alleviating them.