Prediction Sensitivity: Continual Audit of Counterfactual Fairness in Deployed Classifiers
This addresses the challenge of ensuring fairness in AI systems after deployment, particularly for detecting discrimination against individuals, which is critical for high-stakes applications.
The paper tackles the problem of auditing deployed classifiers for individualized fairness by introducing prediction sensitivity, an approach for continual audit of counterfactual fairness, and demonstrates its effectiveness in detecting violations.
As AI-based systems increasingly impact many areas of our lives, auditing these systems for fairness is an increasingly high-stakes problem. Traditional group fairness metrics can miss discrimination against individuals and are difficult to apply after deployment. Counterfactual fairness describes an individualized notion of fairness but is even more challenging to evaluate after deployment. We present prediction sensitivity, an approach for continual audit of counterfactual fairness in deployed classifiers. Prediction sensitivity helps answer the question: would this prediction have been different, if this individual had belonged to a different demographic group -- for every prediction made by the deployed model. Prediction sensitivity can leverage correlations between protected status and other features and does not require protected status information at prediction time. Our empirical results demonstrate that prediction sensitivity is effective for detecting violations of counterfactual fairness.