Path-Specific Counterfactual Fairness
This addresses fairness in AI decision-making for scenarios with complex causal pathways, offering a more generalizable method than previous work.
The paper tackles the problem of learning fair decision systems in complex scenarios where sensitive attributes affect decisions through both fair and unfair pathways, by introducing a causal approach that corrects observations affected by the sensitive attribute to form decisions, avoiding the need for intractable path-specific effect computations.
We consider the problem of learning fair decision systems in complex scenarios in which a sensitive attribute might affect the decision along both fair and unfair pathways. We introduce a causal approach to disregard effects along unfair pathways that simplifies and generalizes previous literature. Our method corrects observations adversely affected by the sensitive attribute, and uses these to form a decision. This avoids disregarding fair information, and does not require an often intractable computation of the path-specific effect. We leverage recent developments in deep learning and approximate inference to achieve a solution that is widely applicable to complex, non-linear scenarios.