Formalising Anti-Discrimination Law in Automated Decision Systems
Not on the map. Why?
Abstract (arXiv)
Algorithmic discrimination is a critical concern as machine learning models are used in high-stakes decision-making in legally protected contexts. Although substantial research on algorithmic bias and discrimination has led to the development of fairness metrics, several critical legal issues remain unaddressed in practice. The paper addresses three key shortcomings in prevailing ML fairness paradigms: (1) the narrow reliance on prediction or outcome disparity as evidence for discrimination, (2) the lack of nuanced evaluation of estimation error and assumptions that the true causal structure and data-generating process are known, and (3) the overwhelming dominance of US-based analyses which has inadvertently fostered some misconceptions regarding lawful modelling practices in other jurisdictions. To address these gaps, we introduce a novel decision-theoretic framework grounded in anti-discrimination law of the United Kingdom, which has global influence and aligns closely with European and Commonwealth legal systems. We propose the "conditional estimation parity" metric, which accounts for estimation error and the underlying data-generating process, aligning with UK legal standards. We apply our formalism to a real-world algorithmic discrimination case, demonstrating how technical and legal reasoning can be aligned to detect and mitigate unlawful discrimination. Our contributions offer actionable, legally-grounded guidance for ML practitioners, policymakers, and legal scholars seeking to develop non-discriminatory automated decision systems that are legally robust.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 47 of its 192 references that are in the map. First posted on arXiv 29 Jun 2024.
Builds on
47 of 192 references.
References · 47
- 2016Equality of Opportunity in Supervised Learningcited by 2,000
- 2019A Survey on Bias and Fairness in Machine Learningcited by 1,911
- 2019Aleatoric and Epistemic Uncertainty in Machine Learning: An Introduction to Concepts and Methodscited by 903
- 2017Counterfactual Fairnesscited by 802
- 2014Certifying and removing disparate impactcited by 715
- 2016Fair prediction with disparate impact: A study of bias in recidivism prediction instrumentscited by 671
- 2017Algorithmic decision making and the cost of fairnesscited by 402
- 2020Closing the AI Accountability Gap: Defining an End-to-End Framework for Internal Algorithmic Auditingcited by 380
- 2020Fairness in Machine Learning: A Surveycited by 369
- 2016Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatmentcited by 368
- 2017Fairness in Criminal Justice Risk Assessments: The State of the Artcited by 315
- 2019Measurement and Fairnesscited by 299
35 more
- 2017Avoiding Discrimination through Causal Reasoningcited by 191
- 2018Path-Specific Counterfactual Fairnesscited by 179
- 2021A survey on datasets for fairness-aware machine learningcited by 157
- 2017Fairness in Machine Learning: Lessons from Political Philosophycited by 148
- 2019On the Apparent Conflict Between Individual and Group Fairnesscited by 126
- 2015Practical Bayesian model evaluation using leave-one-out cross-validation and WAICcited by 110
- 201850 Years of Test (Un)fairness: Lessons for Machine Learningcited by 101
- 2021A Clarification of the Nuances in the Fairness Metrics Landscapecited by 98
- 2018A Unified Approach to Quantifying Algorithmic Unfairness: Measuring Individual & Group Unfairness via Inequality Indicescited by 97
- 2018The Measure and Mismeasure of Fairnesscited by 94
- 2020Why Fairness Cannot Be Automated: Bridging the Gap Between EU Non-Discrimination Law and AIcited by 81
- 2016A causal framework for discovering and removing direct and indirect discriminationcited by 71
- 2019PC-Fairness: A Unified Framework for Measuring Causality-based Fairnesscited by 67
- 2018A Moral Framework for Understanding of Fair ML through Economic Models of Equality of Opportunitycited by 53
- 2020Removing Spurious Features can Hurt Accuracy and Affect Groups Disproportionatelycited by 52
- 2019Discrimination in the Age of Algorithmscited by 51
- 2018Blind Justice: Fairness with Encrypted Sensitive Attributescited by 41
- 2019What does it mean to solve the problem of discrimination in hiring? Social, technical and legal perspectives from the UK on automated hiring systemscited by 38
- 2022Causal Conceptions of Fairness and their Consequencescited by 38
- 2020What's Sex Got To Do With Fair Machine Learning?cited by 37
- 2020On the Moral Justification of Statistical Paritycited by 30
- 2019This Thing Called Fairness: Disciplinary Confusion Realizing a Value in Technologycited by 24
- 2022Marrying Fairness and Explainability in Supervised Learningcited by 23
- 2022The four-fifths rule is not disparate impact: a woeful tale of epistemic trespassing in algorithmic fairnesscited by 21
- 2023Fairness through Aleatoric Uncertaintycited by 16
- 2020Affirmative Algorithms: The Legal Grounds for Fairness as Awarenesscited by 15
- 2020Algorithmic Decision Making with Conditional Fairnesscited by 14
- 2023Algorithmic Unfairness through the Lens of EU Non-Discrimination Law: Or Why the Law is not a Decision Treecited by 14
- 2023Counterfactual Situation Testing: Uncovering Discrimination under Fairness given the Differencecited by 8
- 2022What is Proxy Discrimination?cited by 7
- 2022Transparency, Governance and Regulation of Algorithmic Tools Deployed in the Criminal Justice System: a UK Case Studycited by 7
- 2023Risk Scores, Label Bias, and Everything but the Kitchen Sinkcited by 5
- 2022The Fairness Field Guide: Perspectives from Social and Formal Sciencescited by 3
- 2023On the Cost of Demographic Parity in Influence Maximizationcited by 1
- 2024Hierarchical Bayesian Models to Mitigate Systematic Disparities in Prediction with Proxy Outcomescited by 1
Built on by
Citing papers per year: 2025: 1; 2026: 1.
2 papers.