Algorithmic Fairness: A Runtime Perspective

TL;DR

This paper introduces a framework for analyzing fairness in AI systems as a dynamic, runtime property, emphasizing the importance of sequential monitoring and enforcement in evolving environments.

Contribution

It proposes a minimal model for studying fairness over time, summarizes strategies for monitoring and enforcing fairness, and surveys existing solutions under various environment dynamics.

Findings

Develops a minimal model based on sequences of coin tosses with evolving biases.

Provides general results for fairness monitoring and enforcement under simple assumptions.

Surveys existing solutions for dynamic fairness monitoring and static enforcement.

Abstract

Fairness in AI is traditionally studied as a static property evaluated once, over a fixed dataset. However, real-world AI systems operate sequentially, with outcomes and environments evolving over time. This paper proposes a framework for analysing fairness as a runtime property. Using a minimal yet expressive model based on sequences of coin tosses with possibly evolving biases, we study the problems of monitoring and enforcing fairness expressed in either toss outcomes or coin biases. Since there is no one-size-fits-all solution for either problem, we provide a summary of monitoring and enforcement strategies, parametrised by environment dynamics, prediction horizon, and confidence thresholds. For both problems, we present general results under simple or minimal assumptions. We survey existing solutions for the monitoring problem for Markovian and additive dynamics, and existing solutions for the enforcement problem in static settings with known dynamics.