In search of effectiveness and fairness in proving algorithmic discrimination in EU law
The DOI matches its Crossref record.
Abstract
Examples of discriminatory algorithmic recruitment of workers have triggered a debate on application of the non-discrimination principle in the EU. Algorithms challenge two principles in the system of evidence in EU non-discrimination law. The first is effectiveness, given that due to algorithmic opacity, the parties in algorithmic discrimination cases do not have easy and unrestricted access to facts enabling them to support their claims. The second is fairness, insofar as the launching and unfolding of the evidentiary debate requires lifting the veil of algorithmic opacity: a colossal task, placing unrealistic burdens of proof on claimants and respondents. Algorithmic discrimination thus seems impossible to prove and, consequently, falls outside the scope of application of EU non-discrimination law. Two possible solutions are proposed. First, as regards effectiveness, a right to access evidence in favour of victims of algorithmic discrimination should be recognized, through a joint reading of EU non-discrimination law and the General Data Protection Regulation. Second, to allocate the burden of proof more proportionately, an extension of the grounds for defence of respondents could allow them to establish that biases were autonomously developed by an algorithm.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Recent work has examined evidentiary burdens in proving algorithmic discrimination, pointing to difficulties faced by claimants when access to model logic and performance data is limited (Grozdanovski, 2021).”
From Judicial review of algorithmic administrative systems legality evidence and remedies in the smart city state (2026) · cites this paper as Grozdanovski, 2021
In the citing paper
“Without access to these design layers, claimants face severe barriers in showing how unequal outcomes arise (Grozdanovski, 2021; Wu, 2023).”
From Judicial review of algorithmic administrative systems legality evidence and remedies in the smart city state (2026) · cites this paper as Grozdanovski, 2021
In the citing paper
“EU-focused work argues that existing legal categories do not map cleanly onto how models operate, since the law often expects structured decision logic while algorithmic systems operate through statistical correlations and complex feature spaces (Weerts et al., 2023; Grozdanovski, 2021; Fraenkel-Haeberle, 2026).”
From Judicial review of algorithmic administrative systems legality evidence and remedies in the smart city state (2026) · cites this paper as Grozdanovski, 2021
Citations
Cites
Still being checked.
Cited by
EU Regulation of Artificial Intelligence: Challenges for Patients’ Rights
van Kolfschooten, H. (2022) · Common Market Law Review · doi:10.54648/cola2022005 · Cited by 63
Transparent AI? Navigating Between Rules on Trade Secrets and Access to Information
Mylly, U.-M. (2023) · IIC - International Review of Intellectual Property and Competition Law · doi:10.1007/s40319-023-01328-5 · Cited by 54
Artificial Intelligence and Human Rights
Quintavalla, A., & Temperman, J. (2023) · Oxford University Press eBooks · doi:10.1093/law/9780192882486.001.0001 · Cited by 12
Grozdanovski, L. (2022) · Réseaux · doi:10.3917/res.232.0099 · Cited by 7
Self-citationArtificial Intelligence and Fair Trial Rights
Molbæk-Steensig, H., & Quemy, A. (2023) · Oxford University Press eBooks · doi:10.1093/law/9780192882486.003.0018 · Cited by 10
Metadata
- Authors
- Grozdanovski, L.
- Corresponding author
- Grozdanovski, Ljupcho
- Venue
- Common Market Law Review
- Publisher
- Kluwer Law International BV
- Published
- 2021-01-01
- Volume
- 58
- Issue
- Issue 1
- Pages
- 99-136
- ISSN
- 0165-0750 (ISSN-L)
- OpenAlex
- W3185114540
- Cited by
- 57
- References
- 0
- Citations by year
- 2026: 1 · 2025: 4 · 2024: 2 · 2023: 45 · 2022: 3 · 2021: 2
- FWCI
- 84.71
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Common Market Law Review (publishedVersion)
- Open Repository and Bibliography (University of Liège) (submittedVersion)
- Topics
- Social Sciences › Law › Discrimination and Equality Law
- Social Sciences › Law › Digitalization, Law, and Regulation
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Keywords
- algorithmic discrimination, EU anti-discrimination law, burden of proof, algorithmic opacity, legal effectiveness, fairness, General Data Protection Regulation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.