Fairness as a Program Property

TL;DR

This paper investigates how to verify the fairness of decision-making programs by framing fairness questions as program verification problems and applying automated techniques to assess fairness under probabilistic population models.

Contribution

It introduces a novel approach to verify algorithmic fairness by translating fairness questions into program verification problems and employing automated verification methods.

Findings

Proposes a formal framework for fairness verification.

Demonstrates automated verification techniques for probabilistic models.

Provides insights into the feasibility of verifying fairness in decision programs.

Abstract

We explore the following question: Is a decision-making program fair, for some useful definition of fairness? First, we describe how several algorithmic fairness questions can be phrased as program verification problems. Second, we discuss an automated verification technique for proving or disproving fairness of decision-making programs with respect to a probabilistic model of the population.