Counterfactual Explanations Without Opening the Black Box: Automated Decisions and the GDPR
Open accessThe DOI matches its Crossref record; 7 of the 7 references it cites were found in a registry or an index.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“They are expected to have psychological benefits because they help people act, rather than merely understand, by altering future behavior to achieve a desired outcome (Wachter et al., 2018; Mothilal et al., 2020).”
From Exploring the effects of human-centered AI explanations on trust and reliance (2023) · cites this paper as Wachter et al., 2018
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“On the one hand, 'European conceptualization of discrimination ... is contextual' ([75]) while on the other hand automatic methods to detect and correct discriminatory decisions need clear rules.”
From Fairness and explainability in automatic decision-making systems. A challenge for computer science and law (2023) · cites this paper as [75]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Fifth, even where explanations of AI decisions might be legally required by the GDPR or Convention 108, it is often difficult to explain the logic behind a decision, when an AI system, analysing large amounts of data, arrives at that decision.”
From Discrimination, artificial intelligence, and algorithmic decision-making (2025) · cites this paper as Wachter, Mittelstadt and Russell 2017
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“We show that standard XAI explanations, such as counterfactuals, LIME, or Anchors, even when combined with human intuitions about decision continuity or monotonicity, reveal only errors in the neighborhood of the individual, but provide insufficient grounds for overturning the decision at hand.”
From Explainable AI Isn't Enough! Rethinking Algorithmic Contestability (2026) · cites this paper as Wachter, Mittelstadt, and Russell 2017
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“However, while Wachter, Mittelstadt, and Russell (2017) motivate why counterfactuals can enable individuals to contest decisions and satisfy the legal requirements established by the GDPR, they do not demonstrate how their method can be tailored to this purpose, nor under which conditions contestation based on counterfactuals should be successful.”
From Explainable AI Isn't Enough! Rethinking Algorithmic Contestability (2026) · cites this paper as Wachter, Mittelstadt, and Russell 2017
Experimental — accuracy not yet measured
5 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
A Note on Supra-Syntactic Interference
(1977) · ITL - International Journal of Applied Linguistics · doi:10.1075/itl.37.05mis · Cited by 897
(1974) · The Shock and Vibration Digest · doi:10.1177/058310247400600116 · Cited by 137
(2017) · Acta Ophthalmologica · doi:10.1111/j.1755-3768.2017.02722 · Cited by 46
(2017) · IT Governance Ltd eBooks · doi:10.2307/j.ctt1trkk7x.19 · Cited by 11
ICO report warns pharmacies about data protection breaches
(2017) · Pharmaceutical journal/The pharmaceutical journal · doi:10.1211/pj.2017.20202771 · Cited by 4
(1975) · Brain Research · doi:10.1016/0006-8993(75)90363-7 · Cited by 2
Artikel 13. Informationspflicht bei Erhebung von personenbezogenen Daten bei der betroffenen Person
(2016) · Verlag Dr. Otto Schmidt eBooks · doi:10.9785/9783504384951-054 · Cited by 1
Research works printed but not linked: 26
- E G See, TEXTS, COMMENTARIES AND PRACTICAL GUIDELINES 113
- Heckmann & Paschke, 375, 10.1075/itl.37.05mis · DOI · Paper page
- Rainer Knyrim, 2017, 14 Informationspflicht bei Erhebung von Daten, 417
- Lorenz Franck, 2017, 13 Informationspflicht bei Erhebung von Daten, 338
- 2016, Article 29 Data Protection Working Party, supra note 85, O.J. (L, 101
- Franck, supra note 95, at 322-23. 104. Knyrim, supra note 98, 420
- Id, 2016, at 417; ARTICLE 29 DATA PROTECTION WORKING PARTY, GUIDELINES ON TRANSPARENCY UNDER REGULATION
- The European Commission is tasked in Art. 12(8) to develop such icons
- 2016, See Regulation 2016/679, GDPR, recital 58, O.J. (L, 119
- Heckmann & Paschke, 10.1075/itl.37.05mis · DOI · Paper page
- Franck, at 322 (noting this for the elderly, uneducated people, foreigners, or children); see also Heckmann & Paschke, supra note 92, 376
- Supra Jay, 10.1075/itl.37.05mis · DOI · Paper page
- Heckmann & Paschke, 10.1075/itl.37.05mis · DOI · Paper page
- 2016, Data Protection Working Party, supra note 85, O.J. (L, 110, 41
- Article 29 Data Protection Working Party, supra note 85
- 2018, Note the UK's ICO is preparing new guidelines in the form of a living document, which will be continuously updated. See UK's Information Commissioner's Office, RIGHTS RELATED TO AUTOMATED DECISION MAKING INCLUDING PROFILING
- Info, Comm'r's Office, supra note 112
- Info, Rodway, supra note 95, at 2. 117. Article 29 Data Protection Working Party, supra note 85, at 26. 152. ARTICLE 29 DATA PROTECTION WORKING PARTY, supra note 85, at 26-27. 153. Id. at 27. 154. The intention of Art. 15 is to provide a control mechanism for data subjects to request at any time more or less the same information as Art, 426
- Info, 2016, Franck, supra note 142, at 355; Ehmann, supra note 119, O.J. (L, 156
- 2016, See Regulation 2016/679, GDPR, recital 63 & art, O.J. (L 119), 15
- The European Commission, European Council, and European Parliament in Art
- 2012, (h) proposed to create a legal duty for data controllers to provide any further information beyond those in the notification duties to ensure fair and transparent data processing. See generally EUROPEAN COMMISSION, Regulation of the European Parliament and the Council on the Protection of Individuals with Regard to the Processing of Personal Data and on the Free Movement of Such Data (General Data Protection Regulation
- European, However, this proposal was not adopted and moved to Recital 60, suggesting that there is no legal duty to provide more information than required in Art, 129, 13
- 2016, O.J. (L, 119, 10.2307/j.ctt1trkk7x.19 · DOI
- 2016, L 119) 40 (EU). 205. Regulation 2016/679, GDPR, art, 203. The method described here is compatible with a proposal made by Citron and Pasquale to allow consumers to manually enter "hypothetical alterations" to their credit histories and view their effects. Citron & Pasquale, supra note 15, 679, 46
- 2016, Regulation, 679, 10.2307/j.ctt1trkk7x.19 · DOI
Cited by
Items 1,001–1,053 of 1,053
Challenges and Future Directions
Firoozi, A. A., & Firoozi, A. A. (2024) · Digital innovations in architecture, engineering and construction · doi:10.1007/978-3-031-71097-1_7 · Cited by 1
Computational Argumentation: a Foundation for Human-centric AI
(2024) · Frontiers research topics · doi:10.3389/978-2-8325-5687-0 · Cited by 1
Approaching Principles of XAI: A Systematization
Ronge, R., Bauer, B., & Rathgeber, B. (2024) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2024.3515937 · Cited by 1
The black box and the physician: Investigating the legal issues of explainability in medical AI
Crisp, J., & Desmoulin‐Canselier, S. (2025) · Réseaux · doi:10.3917/res.248.0227 · Cited by 1
The University as an AI Ombudsman—Part 1
Wu, Y. (2025) · Computer · doi:10.1109/mc.2024.3503082 · Cited by 1
Sarkar, S., Das, S., Chanda, A., & Biswas, S. (2025) · Wiley eBooks · doi:10.1002/9781394302444.ch12 · Cited by 1
Learning Feasible Causal Algorithmic Recourse: A Prior Structural Knowledge Free Approach
Wang, H., Zou, H., Zhou, X., Wang, S., Yang, W., & Cui, P. (2025) · International World Wide Web Conference (WWW) · doi:10.1145/3696410.3714859 · Cited by 4
fmeffects: An R Package for Forward Marginal Effects
Löwe, H., Scholbeck, C. A., Heumann, C., Bischl, B., & Casalicchio, G. (2025) · The R Journal · doi:10.32614/rj-2024-024 · Cited by 1
Can Input Attributions Explain Inductive Reasoning in In-Context Learning?
Ye, M., Kuribayashi, T., Kobayashi, G., & Suzuki, J. (2025) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2025.findings-acl.1092 · Cited by 3
Explain it for Safety: Explanations for Risk Mitigation
Schwammberger, M., Rakow, A., Putze, L., & Bairy, A. (2025) · Repository KITopen (Karlsruhe Institute of Technology) · doi:10.5445/ir/1000181811 · Cited by 1
Kar, R. M., & Parycek, P. (2018) · Social Science Open Access Repository (GESIS – Leibniz Institute for the Social Sciences) · Cited by 1
Explaining Machine Learning Models by Generating Counterfactuals
Afonichkin, I. (2019) · Aaltodoc (Aalto University) · Cited by 1
Explanation and Social Scientific Modeling
Vredenburgh, K. (2019) · Digital Access to Scholarship at Harvard (DASH) (Harvard University) · Cited by 1
Lifting Interpretability-Performance Trade-off via Automated Feature EngineeringPreprint
Gosiewska, A., & Biecek, P. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.04267 · Cited by 2
Kemper, C. (2021) · Prometheus · doi:10.13169/prometheus.37.4.0404 · Cited by 1
Algorithm-Agnostic Interpretations for ClusteringPreprint
Scholbeck, C. A., Funk, H., & Casalicchio, G. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2209.10578 · Cited by 1
Unlawful-by-Default:Toward Ex Ante Justification for AiPreprint
Malgieri, G., & Pasquale, F. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4346120 · Cited by 1
Contextual TrustPreprint
Kearns, R. O. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2303.08900 · Cited by 1
Mitigating Spurious Correlations in Machine Learning Models: Techniques and ApplicationsPreprint
Srivastava, M. (2023) · OSF Preprints (OSF Preprints) · doi:10.31219/osf.io/2fsw6 · Cited by 1
Counterfactual explanations for machine learning models on heterogeneous data
Wang, Y. (2023) · DR-NTU (Nanyang Technological University) · doi:10.32657/10356/169968 · Cited by 2
[RETRACTED: The Ethical Implications of Artificial Intelligence on the Human Society]Preprint
Ansari, W. (2023) · doi:10.14293/pr2199.000341.v1 · Cited by 1
Welche Regeln gelten für die Erzeugnisse Künstlicher Intelligenz?
Kreutzer, T. (2021) · doi:10.59350/pvg10-r9d34 · Cited by 1
Pyreal: A Framework for Interpretable ML ExplanationsPreprint
Zytek, A., Wang, W., Liu, D., Berti‐Équille, L., & Veeramachaneni, K. K. (2023) · HAL · doi:10.48550/arxiv.2312.13084 · Cited by 3
Counterfactual Generation with Answer Set ProgrammingPreprint
Dasgupta, S., Shakerin, F., Arias, J., Salazar, E., & Gupta, G. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2402.04382 · Cited by 1
Explainability for Machine Learning Models: From Data Adaptability to User PerceptionPreprint
Delaunay, J. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2402.10888 · Cited by 1
Longitudinal Counterfactuals: Constraints and OpportunitiesPreprint
Asemota, A., & Hooker, G. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2403.00105 · Cited by 1
A.I. and the Nature of DisagreementPreprint
Niblett, A., & Yoon, A. H. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4735851 · Cited by 1
Vial, A. (2022) · theses.fr (ABES) · doi:10.70675/3f561d8bz8e0fz4729za1d8z92b6a6e64298 · Cited by 1
Félix, E., Amadieu, F., & Broisin, J. (2024) · Research Square · doi:10.21203/rs.3.rs-3888008/v1 · Cited by 1
Explaining 3D Computed Tomography Classifiers with CounterfactualsPreprint
Cohen, J., Blankemeier, L., & Chaudhari, A. S. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2502.07156 · Cited by 1
Ethical Considerations in AI Design and Deployment: Insights into Sensitive IndustriesPreprint
Rongali, S. K. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5220284 · Cited by 1
John, A., Hassan, A., & khan, M. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5252622 · Cited by 1
Sutham, N. (2025) · doi:10.58445/rars.2796 · Cited by 1
Optimization by Directional Attacks: Solving Problems with Neural Network SurrogatesPreprint
Bouchet, P.-Y., & Vidal, T. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2510.01461 · Cited by 1
Temraz, M., & Keane, M. T. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7291667 · Cited by 1
The State of Post-Hoc Local XAI Techniques for Image Processing: Challenges and MotivationsPreprint
Poh, R. L. T., Keoh, S. L., & Li, L. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2501.06253 · Cited by 1
An Application of LatentCF++ on Providing Counterfactual Explanations for Fraud Detection
Giannopoulou, M.-S. (2023) · DiVA (Digitala Vetenskapliga Arkivet) · Cited by 2
Counterfactual Explanations on Robust Perceptual GeodesicsPreprint
Zaher, E., Trzaskowski, M., Nguyen, Q. H., & Roosta, F. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2601.18678 · Cited by 2
Local post-hoc interpretability for black-box classifiers
Laugel, T. (2020) · theses.fr (ABES) · doi:10.70675/afa1c80bz1e1fz4930zb0eaz215e7bb5bf65 · Cited by 4
A generic and adaptive approach to explainable AI in autonomic systems : the case of the smart home
Houzé, É. (2022) · theses.fr (ABES) · doi:10.70675/cd5e3060zd7a7z4c5czb731z7a240fad0c49 · Cited by 3
Trustworthy machine learning : explainability and distribution-free uncertainty quantification
Amoukou, S. I. (2023) · theses.fr (ABES) · doi:10.70675/a022ddcbz8047z488bzb33bz68f6b99639b2 · Cited by 6
Interpretable Music Recommender Systems
Afchar, D. (2023) · theses.fr (ABES) · doi:10.70675/acfb6676z8613z4e08zad14z98d4a9c2ee74 · Cited by 3
For an ethics of conversational Artificial Intelligence
Pistilli, G. (2024) · theses.fr (ABES) · doi:10.70675/621c604czc262z4d68z8f9azf5340bfca0ab · Cited by 5
Fischer, S. W., Schraffenberger, H., van Hooff, M. L., Thill, S., & Haselager, P. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.14406 · Cited by 1
Governing Well in the Algorithmic Age: The Foundations of Digital StatecraftPreprint
Engin, Z., Gordon, T., Bastidas, V., Crick, T., Crowcroft, J., Dénis, J.-M., Hand, D. J., Humpherson, E., Maffeo, L., Mökander, J., et al. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.18483 · Cited by 1
CAPTURE: Disentangling Preference Drift from Memory Poisoning in Personalized LLM Agents
Hossain, S. M. A., Shayoni, R. K., & Morol, M. K. (2026) · arXiv · doi:10.48550/arxiv.2609.02265 · Cited by 3
“Instead, for each ledger node n i the agent cites, we re-decode with that node removed and measure the shift in utility, following the counterfactual tradition in explanation (Wachter et al., 2018; Verma et al., 2024).”
Explainability in AI: Comparing Human-Like and System-Like Trust Repair Strategies
Konopka, B., & Wiesche, M. (2026) · Information Systems Frontiers · doi:10.1007/s10796-026-10751-1 · Cited by 1
احريش, ع. (2026) · PRSM · doi:10.34874/prsm/contentieux-affa-70387
L'acte administratif algorithmique
Stepanov, A. (2026) · doi:10.70675/adc05be3za62ez4b4fz81e7zabf744e60dbf
Human-Clinical AI Agent Collaboration
Kadem, M., & Al-Khazraji, B. (2025) · Proceedings of the AAAI Symposium Series · doi:10.1609/aaaiss.v6i1.36060
Designing “Contestability” in Automated HR Decisions: Appeal Mechanisms, Evidence, and Outcomes
Professor, S. O. C., Radha, P., Chandraseka, V., Associate Professor, S. O. C., Vennila, R., & Professor, S. O. C. (2025) · GS WOW: Wisdom of Worthy Research Journal · doi:10.62078/grks.2025.v04i02.002
Explainable AI Isn't Enough! Rethinking Algorithmic Contestability
Freiesleben, T., Meding, K., & König, G. (2026) · arXiv · doi:10.48550/arxiv.2605.16041
Fraser, H. L., & Stardust, Z. (2026) · Law & Society Review · doi:10.1017/lsr.2026.10124
Metadata
- Authors
- Wachter, S., Mittelstadt, B., & Russell, C.
- Affiliations
- University of Oxford (GB)
- Corresponding author
- Wachter, Sandra
- Venue
- SSRN Electronic Journal
- Publisher
- Elsevier BV
- Published
- 2017
- Volume
- 31
- Issue
- 2
- Pages
- 841
- ISSN
- 1556-5068 (ISSN-L)
- OpenAlex
- W2765204106
- Cited by
- 1,061
- References
- 7
- Citations by year
- 2026: 82 · 2025: 166 · 2024: 192 · 2023: 200 · 2022: 163 · 2021: 143 · 2020: 75 · 2019: 28 · 2018: 11 · 2017: 1
- Type
- posted content
- Language
- en
- Open access
- green
- doi.org
- Locations
- SSRN Electronic Journal (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Alan Turing Institute (EP/N510129/1., N510129)
- Engineering and Physical Sciences Research Council (N510129, EP/N510129/, EP/N510129/1)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Social Sciences › Law › Digitalization, Law, and Regulation
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Keywords
- counterfactual explanations, right to explanation, GDPR, automated decision-making, algorithmic transparency, black box, algorithmic accountability, data subjects, data controllers, trade secrets, privacy, interpretability
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.