​​Artificial Intelligence Predictive Modelling for Forced Displacement Preparedness   

Overall objectives

The main objective of this activity is to strengthen Uganda’s preparedness for large‑scale refugee inflows by operationalizing an Artificial Intelligence model capable of predicting forced displacement movements from the Democratic Republic of the Congo (DRC) and South Sudan. By producing reliable, near real‑time forecasts, the activity aims to strengthen national preparedness by enabling anticipatory action and reducing pressure on already strained hosting districts. 

The activity will deliver actionable predictions that can be directly incorporated into the Government of Uganda’s planning and resource‑allocation processes. It will also build national capacity to use, manage and interpret AI‑generated forecasts, ensuring that government officials, humanitarian actors and development partners can apply these insights to improve readiness, allocate resources more efficiently and protect both refugees and host communities. 

Through these outcomes, the activity seeks to enhance Uganda’s ability to respond proactively to displacement dynamics, support the sustainability of its refugee‑inclusive policy framework and reduce the human and financial costs of reactive, post‑arrival interventions. 

Activity description

Building on analytical work completed in 2025 by the World Bank, this activity will operationalize an AI‑based forecasting model designed to predict refugee inflows into Uganda from the Democratic Republic of the Congo and South Sudan. After reviewing the prototype, the Government of Uganda requested support to render the model fully operational, integrate its predictions into planning processes and strengthen national capacity to use the tool. 

The activity will be implemented through three components: 

Operationalization of the forecasting model: The model will be updated with recent data and run in near real time, producing bi‑weekly predictions accessible through an online platform to support anticipatory action. 

Capacity‑building: Targeted training for government, humanitarian and development actors that will be delivered through a mix of in‑person and virtual sessions that will enable stakeholders to interpret and apply model outputs in policy, budgeting and operational planning. 

Dissemination and learning: A structured outreach strategy, including dedicated events and engagement with a broad range of partners, will promote understanding and uptake of AI‑generated forecasts across Uganda’s refugee response system. 

Engagement with partners

The design and implementation of this activity will rely on close collaboration between the World Bank and UNHCR, supported by targeted engagement with key stakeholders across government, humanitarian and development partners. The World Bank will update and operationalize the forecasting model, support capacity building activities and coordinate analytical and dissemination work. UNHCR will contribute its expertise on displacement dynamics, provide operational context and facilitate engagement with partners in Uganda, regional and headquarters teams. 

Additionally, this activity will engage the Government of Uganda, particularly the Office of the Prime Minister–Department of Refugees and the Ministry of Finance, Planning and Economic Development to ensure that the model’s predictions are relevant for planning and aligned with national priorities. Regular consultations with humanitarian agencies, development partners and academic institutions will provide tailored inputs on model use, support capacity-building efforts and encourage uptake of AI‑generated forecasts across policy, programming and operational decision‑making.