184 THE LAST WORD Over the last decade, the number of forcibly displaced people worldwide has almost doubled, reaching 123.2 million in 2024. Managing and supporting displaced populations has become a critical responsibility for governments and humanitarian actors globally. Climate change is also driving displacement, with the number of climate migrants expected to rise to between 25 million and one billion by 2050. Through recent refugee crises, including in Ukraine, we have witnessed the increasing role technology plays in the daily lives of migrants. For example, in Poland the government extended its national e-identification database to include refugees to ensure access to public services and resources, while the World Food Program rolled out digital cash assistance programmes in partnership with a Ukranian bank. Many of the latest innovations in humanitarian technology are being propelled by advances in AI, including initiatives that support governments to prepare for refugee movements before they happen. For example, the World Bank has developed an AI-powered refugee forecasting tool to predict refugee arrivals, drawing on over 90 variables such as conflict, economic trends, climate data, news and social media. The model has been tested in Uganda to predict the scale, location and timing of displacement. Going forward, the tool will help humanitarian actors begin preparations up to six months in advance. Similarly, the United Nations High Commissioner for Refugees (UNHCR) and its partners piloted a project using computer vision and satellite imagery of Somalia to detect drought and correlate this with conflict and displacement patterns. These initiatives support a shift from ad-hoc crisis management to proactive planning, preventing delays and supporting the effective mobilisation of critical resources. The application of AI also raises risks and controversies. The Migrants Rights Network and other non-government organisations are closely monitoring the integration of AI into government immigration and asylum systems. Data ethics, transparency and privacy are particularly important when working with vulnerable populations and sensitive personal data. International guidelines, such as UNHCR’s own Information Integrity Toolkit, include specific provisions on AI and ethics, and strongly emphasise human oversight, algorithmic accountability and the principle of ‘do no harm’. As displacement crises become more frequent and complex, the question isn’t whether AI will shape humanitarian response, but whether we can harness its predictive power while preserving human dignity. The early evidence suggests that when designed with transparency, accountability and refugee agency at its core, AI can transform our capacity to prepare for – rather than merely react to – the movements of the world’s most vulnerable people. Louisa Barker is a senior research manager for IDC Government Insights AI is helping governments and humanitarian actors prepare for displacement before it happens LOUISA BARKER: IDC From reactive to predictive
RkJQdWJsaXNoZXIy NzQ1NTk=