CONCEPT

Relational Ethics (Birhane)

Abeba Birhane’s framework for AI ethics that begins from a relational ontology—the proposition that persons are not bounded data points but nodes in dense webs of relation—and that therefore judges any AI system not by its technical accuracy but by whether it respects the irreducible, co-constituted humanity of those it acts upon.

Relational ethics is Abeba Birhane's alternative to the dominant paradigm of AI ethics, which she argues is built on a false model of the human. The dominant paradigm treats persons as discrete, stable, fully describable units that can be measured, modeled, and predicted. It responds to harms with technical fixes—debiasing algorithms, fairness metrics, transparency tools—that leave the underlying model of the human intact. Birhane's framework begins one level deeper, from what she calls a fundamental shift from the rational to the relational: the proposition, drawn from embodied cognitive science and Sub-Saharan African relational philosophies, that existence is fundamentally co-existent in a web of relations. A person is not a bounded data point but a node in a dense net of relationships, contexts, and histories, continually constituted and reconstituted through interaction with others. The self is relational all the way down. This ontological starting

Keep reading with YOU ON AI