Challenges of Direct Discrimination in Algorithmic Recruitment: Insuperable or Not?*

Abstract

Although the EU prohibition of direct and indirect discrimination under non-discrimination law theoretically applies to algorithmic recruitment, doubts exist about whether the prohibition can tackle algorithmic recruitment discrimination in practical terms. This article examines two significant obstacles to dealing with direct discrimination identified in prior research: ambiguity in determining whether proxy discrimination constitutes direct discrimination, and the incapacity of job applicants to meet the burden of proof. First, the analysis of the Court of Justice of the European Union (CJEU) case law suggests that the inextricable link doctrine could apply to proxies used by algorithmic recruitment systems, assuming that the proxies are comprehensible and detectable. In cases in which the proxies remain unintelligible or invisible, it might still be possible to show that the protected ground determined the decision to impose less favourable treatment. Thus, proxy discrimination in algorithmic systems could constitute direct discrimination. Moreover, the direct discrimination prohibition could deal with more proxy discrimination cases if the applicability of the Charter of Fundamental Rights of the European Union (CFREU) is acknowledged when an EU Regulation directly applies to algorithmic recruitment. Second, the research discloses that the burden of proof could be interpreted in a way which allows applicants to establish prima facie cases of algorithmic recruitment discrimination even if they lack access to detailed information about the workings of the algorithmic recruitment system. The challenges of direct discrimination in algorithmic recruitment might not be insuperable, and the prohibition of direct discrimination should not be sidestepped when analysing the discriminatory tendencies of algorithmic recruitment.