The project advanced the philosophical foundations of algorithmic fairness in several directions. It showed that standard group fairness criteria — statistical parity, calibration, equalised odds — cannot be straightforwardly derived from established theories of distributive justice. This result, established formally in a mathematical proof, reframes the debate: rather than asking which fairness metric is “correct,” the question becomes what moral principles justify choosing one criterion over another in a given decision context.

The principle of fair equality of chances, developed across several of the project’s publications, offers one such justification. Applied to healthcare, it provides a framework for evaluating when prediction-based triage, screening, or resource allocation treats patients fairly — not merely accurately. The subsequent work on counterfactual fairness and causal equal protection extended this line of inquiry, connecting the philosophical analysis to the technical machinery of causal inference.

At Politecnico di Milano, the project created a productive bridge between the mathematics department and the philosophy of science group, contributing to a research environment where formal methods and normative reasoning inform each other. Several of the collaborations initiated during the fellowship — with Nicolò Cangiotti, Marcello Di Bello, Francesco Nappo, and Clinton Castro — continue to produce joint work.