The earlier humanitarian research supplies questions about useful tasks, organisational readiness and consequences for affected people. The DCI/GIZ publications provide the classification and risk assessment foundations for this adaptation. This is the proposed relationship between the works, not a claim that one publication formally derived from another.
Earlier humanitarian work
The NRC Ukraine report was completed in April 2024 and published in October 2024. It examines organisational uses and distinguishes direct prompting from integrated systems. The May 2024 cash-assistance report discusses opportunities across the programme cycle, alongside risks and adoption barriers.
What the new work adds
The taxonomy turns these questions into a consistent record of a defined use. Keep what a system can do separate from what it is permitted to do. Link proposed benefits to evidence, and record when a person must intervene.
What needs fresh evidence
Historical product descriptions, benefit claims and legal statements need checking before operational reuse. A proposed application is not evidence of deployment or measured impact. The earlier publications are part of the research history, not endorsements of this collection.
Sources
Responsible AI Adoption in Complex Environments: An Analysis for NRC Ukraine’s Humanitarian Efforts
A. A Taxonomy for AI in Social Protection: Defining Capabilities and Application Areas