This PRIO Paper aims to take stock of how we are having AI ethics conversations in the aid sector. Recognition of current and future ‘humanitarian AI problems’ gives rise to a ‘lack-of-ethics narrative’. The point of departure for this paper is that the myriads of ongoing ethics conversations are too similar and too open-ended. This need not continue to be the state of how we do this. I suggest that we grapple more carefully and systematically with the questions that arise before looking for answers. In short: what do humanitarians talk about when they talk about humanitarian AI problems needing ethics? Should there be such a thing as humanitarian AI ethics, and if so, for what? This paper provides structure to these conversations by identifying and unpacking four frames underpinning them. The frames are perspectives on technology, AI imaginaries, experimentation, and ethics. The intervention is directed at practitioners, policymakers, and academics alike.
Sandvik, Kristin Bergtora (2024) Framing Humanitarian AI Conversations: What Do We Talk About When We Talk About Ethics?. PRIO Paper. Oslo: PRIO.