Operationalizing Reparative Algorithmic Impact Assessments: Four Technical Artifacts for Healthcare AI Governance

Algorithmic impact assessments (AIAs) and fairness audits, as currently practiced, miss dimensions that prove central to AI deployments in lowresource healthcare contexts, particularly across the Global Majority. These include power asymmetries between developers, institutions, and affected communities; the political economy and sovereignty conditions of training data; the epistemic exclusion of non-Western knowledge systems; and the distribution of technical capacity between deployment partners as systems persist over time. This paper specifies a Reparative Algorithmic Impact Assessment (R-AIA) protocol that augments standard AIA tooling along these dimensions. Building on prior work (Racine, 2024; 2025), the protocol pairs a six-step evaluation framework with four classes of technical artifact: provenance audit trails, co-constructed impact taxonomies, community-defined evaluation rubrics, and capacity-transfer metrics. For the provenance artifact we give a concrete relational schema, example records, and an audit query. For the others, we provide structured specifications and worked examples. We then illustrate the protocol through a hypothetical maternal health system in Sub-Saharan Africa, analyzing its implementation burden, failure modes, and capture risks in the process. Taken together, we strive to offer a minimum viable instantiation for resource constrained settings. The R-AIA gives technical AI governance a concrete instrument for assessing the structural conditions that determine whether such deployments serve or harm the populations they target.