Humanitarian Group Embraces AI to Reach Millions in Crisis Zones

The International Rescue Committee is moving to integrate artificial intelligence into its humanitarian operations, arguing that organizations serving people in conflict and crisis should be among the technology's most active users rather than only its skeptics. The group's research and innovation lead said the conclusion followed years of testing AI in refugee camps, conflict zones and disaster responses.

The push comes at a moment of severe strain for humanitarian aid. More conflicts are underway than at any time since the Second World War, climate disasters are accelerating, and almost 260 million people need assistance, including more than 118 million forcibly displaced — a record. Governments are retreating from funding, and aid has contracted 25% since 2024. Clinics are closing, schools are shutting, and vaccines are not being delivered. Hundreds of thousands of preventable deaths, if not millions, are expected between now and 2030.

That gap is the moral case for AI in humanitarian settings, the IRC argues, and also the reason the technology cannot be deployed carelessly. People in crisis have the least power to refuse a technology, contest its errors or walk away from it, so they should never serve as anyone's test case. The question was never whether AI could help, but what it takes to find out safely.

Several programs illustrate the approach. Signpost, the organization's digital information service, has reached 20 million people, nearly 591,000 of them through one-to-one guidance from moderators, many of whom are displaced themselves. An AI mentor called aprendIA reaches teachers through WhatsApp, the technology they already use. In Nigeria it is on track to reach nearly a million children by the end of 2026 and 6.9 million by 2028, at an estimated $0.20 per child.

The IRC is also piloting AI and satellite imagery to locate children who never appeared in census records as conflict moves entire communities. Without knowing where children are, vaccinating them is impossible. By adapting delivery routes as climate and conflict conditions change, the tool could cut the cost per child by 25% while surfacing 10% to 20% more children. Pilots launch in Somalia next quarter.

Globally, 43 million children under five suffer acute malnutrition. The organization's simplified treatment programs work and are 20% cheaper than standard mechanisms, but manual, error-prone data collection holds them back. It is testing photo-to-digital capture to replace hand transcription, flag children who drop out of care, and read rainfall and displacement signals to act before caseloads surge.

Humanitarian workers, like doctors, are guided by the principle of doing no harm. When mistakes can cost lives, moving fast and breaking things is not a strategy. The IRC built a prototyping and evaluation capability whose job is not to launch AI but to break it first, creating a resource the whole organization draws on.

Before Signpost's AI assistant went near a client, it ran for six months across Greece, Italy, El Salvador and Kenya while moderators and protection officers scored its answers against three standards: whether they were safe, client-centered and trauma-informed. The system's pass rate climbed from 52% to 77%, and staff reported working about 70% faster. But it still failed roughly one answer in four, so it was not deployed on its own. Every response went through a human, and the most sensitive questions never reached the model at all.

That discipline extends to the rest of the work. The organization tracks guidance from data protection and cybersecurity agencies, participates in peer networks where security professionals share emerging threats in real time, reviews use cases carefully and revisits its policies continuously. It designs for low bandwidth, old phones and dozens of languages, and builds nothing a human cannot audit, override or switch off.

AI cannot replace a caseworker with lived experience, and the IRC says it is not trying to. The goal is to let teams reach far more people. Because the organization works where the stakes are highest and the margin for error smallest, its applications could help set guardrails for the wider industry. Staying away, it argues, would mean accepting that people living through war, displacement and climate disaster become the last to benefit from technologies everyone else enjoys.

5Views

Author

Caroline Mercer

World News Correspondent

Caroline Mercer covers public affairs, politics, business, culture and daily news for Core Memo. The role focuses on verification, context, and clear explanations for readers.

All articles by this author