I’m a humanitarian leader who’s all in on AI. Here’s why

Humanitarian leader Jennifer states that the International Rescue Committee (IRC), where she leads research and innovation, is embracing artificial intelligence (AI) to serve people caught up in conflicts and crises. After rigorous testing of AI in various challenging environments, Jennifer believes that humanitarian organizations should not only be skeptics of AI but also its strongest advocates, users, and learners.

This shift is necessary due to the increasing number of conflicts and climate disasters, which have left millions in need of assistance.

The IRC has successfully implemented AI in refugee camps, conflict zones, and disaster responses, leading to significant improvements in service delivery. For instance, their digital information service, Signpost, has reached 20 million people, with nearly 591,000 receiving personalized guidance from moderators. An AI mentor named aprendIA is being used in Nigeria to reach nearly a million children by the end of 2026 and 6.9 million by 2028, at a cost of $0.20 per child.

Additionally, AI is being piloted to locate children who missed census records, crucial for vaccine distribution in war-torn regions.

Jennifer emphasizes that while AI has potential risks such as bias, hallucinations, and exposure of sensitive information, these risks can be mitigated through responsible use. To ensure safety, the IRC built a prototyping and evaluation capability that tests AI extensively before deployment. They conducted rigorous evaluations of an AI assistant used in Greece, Italy, El Salvador, and Kenya, which improved its performance from a 52% pass rate to 77%.

However, the system still fails approximately one in four responses, leading to a cautious approach where human oversight is necessary for sensitive questions.

Jennifer stresses the importance of adhering to the principle "do no harm" in humanitarian settings, where mistakes can have dire consequences. The IRC's responsible AI approach involves continuous policy review, adherence to data protection and cybersecurity standards, and building solutions adaptable to low-bandwidth environments and multiple languages.

By leveraging AI carefully and transparently, Jennifer believes the humanitarian sector can lead the way in setting industry-wide guardrails, ensuring that technological advancements benefit those most in need.