Artificial Intelligence (AI) is reshaping the humanitarian sector in fundamental ways. Donors and the humanitarian sector are grappling with how to understand the nature of this transformation. The promise generative AI (GenAI) comes with in a sector where time and resources are always scarce is salient. Yet, in a sector that serves individuals in precarious and vulnerable situations, relying on AI-generated information may be fraught with mistakes and pitfalls.
Whereas important scholarly and practitioner focus has been given to shifts in risk and needs assessment and aid delivery, and issues of humanitarian accountability and governance, much of this amount to stock taking exercises. Taking this emergent scholarship and ongoing policy initiatives as its point of departure, KnowingAID takes a step back to question how GenAI affects the production of knowledge – through numbers, text, imageries and ultimately discourses – in aid.
Akin to previous phases of the digital transformation of aid, current discussions on AI vacillate between techno-optimism and skepticism regarding the expectations and concerns about how technology can both improve and jeopardize aid and expose recipients and humanitarian workers to new types of risk and harm. Yet, underpinning these discussions is a fundamental shift in how knowledge about humanitarian situations is produced, disseminated and adopted, along with a concern about how the sector should engage.
This Strategic Initiative sets the stage for future investigations into the following questions: how will AI shape (1) the knowledge production on ongoing humanitarian situations; (2) the design and implementation of responses; and (3) humanitarianism as a cosmopolitan idea and practice?
Our approach: We propose to study these questions through three levels, and how GenAI plays into: (1) the creation of data (numbers), text and images; (2) needs assessment and relief distribution and (3) communication about crises.
