Datafied localization: Reproducing unequal power hierarchies in humanitarianism

The integration of artificial intelligence, machine learning, and algorithms into development aid and relief efforts has brought datafication processes to the forefront of contemporary humanitarianism. This article contributes to critical studies of (big) data in humanitarianism by introducing insights from the humanitarian literature on distance and proximity. It highlights three ways in which the mobilization of big data to convey “local” perceptions of needs—what we call datafied localization —can perpetuate epistemic injustices and paternalistic logics in humanitarian practice. First, we argue that big data represents a recontextualization rather than a decontextualization of reality, underscoring that data only becomes knowledge through contextualized interpretation. Second, we propose that the question of defining the local becomes even more pertinent in a datafied version of reality as representation takes place at a distance. Third, we argue that the introduction of big data into localization efforts risks reducing local communities to mere data suppliers, thereby reproducing the power imbalances that localization was meant to counter.