Why Fair Automated Hiring Systems Breach EU Non-Discrimination Law

内容摘要

Employment selection processes that use automated hiring systems based on machine learning are becoming increasingly commonplace. Meanwhile, concerns about algorithmic direct and indirect discrimination that result from such systems are front-and-center, and the technical solutions provided by the research community often systematically deviate from the principle of equal treatment to combat disparate or adverse impacts on groups based on protected attributes. Those technical solutions are now being used in commercially available automated hiring systems, potentially engaging in real-world discrimination. Algorithmic fairness and algorithmic non-discrimination are not the same. This article examines a conflict between the two: whether such hiring systems are compliant with EU non-discrimination law.