Humanitarian Groups Use AI to Speed Disaster Relief Efforts

Humanitarian organizations are increasingly adopting artificial intelligence (AI) to expedite their response to natural disasters and other crises, aiming to reach affected populations faster and more effectively. This integration of AI is particularly crucial as the humanitarian sector faces significant funding shortfalls, estimated in the billions of dollars. GiveDirectly, for instance, utilized an internal AI tool to generate an immediate snapshot of a disaster's impact following earthquakes in Venezuela in June. Subsequently, the organization employed AI-driven satellite imagery analysis to pinpoint the most severely affected neighborhoods. This data was then cross-referenced with poverty metrics, leading to the distribution of flyers inviting families to apply for aid. Within three weeks, GiveDirectly began disbursing cash directly to the mobile phones of individuals in need, demonstrating a faster aid delivery mechanism compared to some traditional methods. The application of AI in these scenarios often involves a collaborative approach between technology and human oversight. While AI tools excel at identifying damage from satellite imagery, their accuracy is not yet absolute. GiveDirectly, therefore, dispatched a ground team in Venezuela to verify the extent of damage and prioritize areas for assistance. Similarly, an AI tool was used to identify emerging tent communities post-disaster, with human volunteers from Humanitarian OpenStreetMap Team (HOT) verifying these AI-generated results. Mercy Corps is employing agentic AI to enhance the speed and efficiency of report generation for its analysts. Nayid Orozco, AI solutions and delivery manager at Mercy Corps, stated that the AI "doesn’t replace our analysts" but rather handles "mechanical work or repetitive research." This allows analysts to dedicate more time to leveraging their expertise and making critical decisions. In Sudan, Mercy Corps recently deployed an internal tool, developed in collaboration with Cloudera, to aggregate security data, displacement figures, and other reports, including news articles in multiple languages. Previously, analysts would spend several hours to days gathering such information. The AI tool, which is designed to use only specific, trusted sources and provides citations, has effectively halved the time required to produce these comprehensive reports. The International Rescue Committee (IRC) is also exploring AI for similar purposes, aiming to improve operational efficiency and the speed at which critical information is processed and acted upon in complex emergency environments. The broader trend indicates a strategic shift towards integrating AI to overcome logistical hurdles and resource constraints, thereby maximizing the reach and impact of humanitarian aid.