New brief: Artificial intelligence in the humanitarian sector

At The MERL Tech Initiative, we think creating accessible information around technological advancements is critical, especially for those making decisions that can impact individuals in highly vulnerable circumstances. We are excited to share our humanitarian AI brief produced as part of ALNAP’s “Explain Briefs“, a series of essential briefings for humanitarian decision makers. 

In this new publication, we are synthesising key information about AI’s varied applications for critical humanitarian decision-makers. Designed to provide a usable baseline for those interested in humanitarian AI and trying to learn more about a rapidly changing context, the brief includes an overview of AI applications and current use across 4 areas:

• Efficiency gains in everyday tasks
• Augmenting institutional workflows
• Enhancing community engagement
• Expanding humanitarian capabilities

“When developing AI guidelines, humanitarians must consider the full spectrum of AI tools and their functions. Clearly defining where and how these tools are used in humanitarian programming helps assess their risks, benefits, and relevance in context.”

The brief focuses on evidence and knowledge regarding AI use, and emphasises the need for more shared learning and humanitarian specific research on the realised benefits of humanitarian AI as compared to the costs of developing, testing, training and ensuring ongoing safe and responsible use of AI. We offer key considerations as humanitarians navigate this new landscape and hope that humanitarian decision makers dedicate the resources necessary to help develop a more discerning and detailed approach to the use of AI in the sector. 

Crucial to developing this responsible, coherent approach is the process of critically assessing claims about GenAI’s potential to enhance efficiency, while weighing the risks and broader implications—especially for crisis-affected communities. In the brief we caution rushing to make decisions and suggest pathways for the humanitarian space to develop a more purposeful approach to AI that separates hypothetical potential from proven impact. 

We recognise that things are changing quickly and are keen on offering evergreen guidance with timely knowledge, so we hope this brief will be a resource that provides relevant insight and guidance for those seeking to make informed decisions regarding the use of AI in the humanitarian sector.

You can read the brief here, and if you’re pressed for time a short version of the brief here.