Auditing for Bias in Ad Delivery Using Inferred Demographic Attributes

TL;DR

This paper examines how demographic inference errors affect bias auditing in ad delivery and proposes a method to adjust for these errors, improving the accuracy of bias detection in black-box settings.

Contribution

It introduces a novel adjustment technique to mitigate inference errors in demographic-based bias audits of ad delivery algorithms.

Findings

Inference errors can cause false negatives in bias detection.

Adjusting for expected inference error improves bias detection sensitivity.

First study to address inference error impact in black-box ad bias auditing.

Abstract

Auditing social-media algorithms has become a focus of public-interest research and policymaking to ensure their fairness across demographic groups such as race, age, and gender in consequential domains such as the presentation of employment opportunities. However, such demographic attributes are often unavailable to auditors and platforms. When demographics data is unavailable, auditors commonly infer them from other available information. In this work, we study the effects of inference error on auditing for bias in one prominent application: black-box audit of ad delivery using paired ads. We show that inference error, if not accounted for, causes auditing to falsely miss skew that exists. We then propose a way to mitigate the inference error when evaluating skew in ad delivery algorithms. Our method works by adjusting for expected error due to demographic inference, and it makes skew detection more sensitive when attributes must be inferred. Because inference is increasingly used for auditing, our results provide an important addition to the auditing toolbox to promote correct audits of ad delivery algorithms for bias. While the impact of attribute inference on accuracy has been studied in other domains, our work is the first to consider it for black-box evaluation of ad delivery bias, when only aggregate data is available to the auditor.