The DOI matches its Crossref record; 20 of the 20 references it cites were found in a registry or an index.
Demystifying the Draft EU Artificial Intelligence Act
Open accessPreprint — published as 10.9785/cri-2021-220402 · Paper page
Abstract
In April 2021, the European Commission proposed a Regulation on Artificial Intelligence, known as the AI Act. We present an overview of the Act and analyse its implications, drawing on scholarship ranging from the study of contemporary AI practices to the structure of EU product safety regimes over the last four decades. Aspects of the AI Act, such as different rules for different risk-levels of AI, make sense. But we also find that some provisions of the draft AI Act have surprising legal implications, whilst others may be largely ineffective at achieving their stated goals. Several overarching aspects, including the enforcement regime and the effect of maximum harmonisation on the space for AI policy more generally, engender significant concern. These issues should be addressed as a priority in the legislative process.
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 cited source says what the claim attributes to it.
In the citing paper
“The legislation will eventually be complemented by technical standards developed by European standardization organizations to at least partially remove this legal uncertainty and to facilitate the implementation of requirements for high-risk systems, including human oversight [86].”
From On the Quest for Effectiveness in Human Oversight: Interdisciplinary Perspectives (2024) · cites this paper as [86]
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
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
Kaugeranna, R., Kaugeranna, E., & Aion, (. S. 4. 6. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.giscience.2023.43 · Cited by 13,690
Emotional expressions reconsidered: Challenges to inferring emotion from human facial movements
Barrett, L. F., Adolphs, R., Marsella, S., Martinez, A. M., & Pollak, S. D. (2019) · Psychological Science in the Public Interest · doi:10.1177/1529100619832930 · Cited by 1,748
Unpacking the Social Media Bot: A Typology to Guide Research and Policy
Gorwa, R., & Guilbeault, D. (2020) · Policy & Internet · doi:10.1002/poi3.184 · Cited by 270
The Ethics of Emotion in Artificial Intelligence Systems
Stark, L., & Hoey, J. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445939 · Cited by 179
Platform values and democratic elections: How can the law regulate digital disinformation?
Marsden, C., Meyer, T., & Brown, I. (2020) · Computer Law & Security Review · doi:10.1016/j.clsr.2019.105373 · Cited by 161
X-LXMERT: Paint, Caption and Answer Questions with Multi-Modal TransformersPreprint
Cho, J., Lu, J., Schwenk, D., Hajishirzi, H., & Kembhavi, A. (2020) · Conference on Empirical Methods in Natural Language Processing · doi:10.48550/arxiv.2009.11278 · Cited by 109
X-LXMERT: Paint, Caption and Answer Questions with Multi-Modal Transformers
Cho, J., Lu, J., Schwenk, D., Hajishirzi, H., & Kembhavi, A. (2020) · Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2020.emnlp-main.707 · Cited by 99
ShaRF: Shape-conditioned Radiance Fields from a Single ViewPreprint
Rematas, K., Martin-Brualla, R., & Ferrari, V. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.08860 · Cited by 90
Theorizing Regulatory Intermediaries: The RIT ModelPreprint
Abbott, K. W. (2017) · SSRN Electronic Journal · doi:10.2139/ssrn.2925411 · Cited by 67
Research works printed but not linked: 8
- Marijn Sax, 'Between Empowerment and Manipulation: The Ethics and Regulation of For-Profit Health Apps' (PhD Thesis, Univeristeit van Am- sterdam (UvA) 2021) 110-12.
- See e.g. Oscar H Gandy, Coming to Terms with Chance: Engaging Ra- tional Discrimination and Cumulative Disadvantage (Routledge 2009).
- See generally Nick Seaver, 'Captivating Algorithms: Recommender Sys- tems as Traps' (2019) 24 Journal of Material Culture 421. Harms are identified especially in relation to children, see e.g. Beeban Kidron and others, 'The Cost of Persuasive Design' (5 Rights Foundation, June 2018) https://5rightsfoundation.com/uploads/5rights-disrupted-childhood-digit al-version.pdf.
- See generally Evan Stark and Marianne Hester, 'Coercive Control: Update and Review' (2019) 25 Violence Against Women 81 (on how the concept of coercive control entered English law due to how 'discrete, injurious as- saults [were] too narrow to capture [patterns] of coercion').
- Jevan A Hutson and others, 'Debiasing Desire: Addressing Bias & Discri- mination on Intimate Platforms' (2018) 2 Proc ACM Hum-Comput In- teract 73:1;
- See, on multi-stage profiling, Reuben Binns and Michael Veale, 'Is That Your Final Decision? Multi-Stage Profiling, Selective Effects and Arti- cle 22 of the GDPR' (Presented at PLSC-EU 2019, on file with authors 2021). Classification itself may be considered a form of harm (of repre- sentation), but there is little legal protection around this. See Reuben Binns, 'Fairness in Machine Learning: Lessons from Political Philosophy' (2018) Conference on Fairness, Accountability and Transparency (FAT* 2018), 8.
- Stephen Weatherill, 'The Fundamental Question of Minimum or Maxi- mum Harmonisation' in Sacha Garben and Inge Govaere (eds), The In- ternal Market 2.0 (Hart Publishing 2020). 151 AI Act, art 1(1). 152 AI Act, art 3(1). 153 Decision 768/2008/EC, Annex II. 154 The remaining areas are left to TFEU 34 and 36 to govern what remain- ing restrictions are permitted. 155 With the exception of AI systems developed or used exclusively for mili- tary purposes, and, to the criticism of the EDPS and EDPB, to interna- tional law enforcement co-operation, which they see as a loophole. See European Data Protection Board and European Data Protection Supervi- sor (n 47) para 14. 156 See generally Roel Dobbe and Meredith Whittaker, 'AI and Climate Change: How They're Connected, and What We Can Do about It' (AI Now Institute, 17 October 2019) https://medium.com/@AINowInstitute/a i-and-climate-change-how-theyre-connected-and-what-we-can-do-about -it-6aa8d0f5b32c accessed 2 July 2021. 157 For example, Member States may still be able to, with the permission of the Commission, introduce a measure relating to 'protection of the envir- onment or the working environment on grounds of a problem specific to that Member State' on the basis of scientific evidence, or 'public health' despite maximum harmonisation, see TFEU, arts 114(5), 114(8). 158 The Commission notes that '[the end-user is not one of the economic operators who bear responsibilities under Union harmonisation legisla- tion' in relation to 'any operation or transaction', although this might 'fall under another regulatory regime, in particular at national level.' See Commission Notice of the 27th July 2016 on the 'Blue Guide' on the im- plementation of EU products rules 2016, OJ C272/1
- Veale / Zuiderveen Borgesius -Demystifying the Draft EU Artificial Intelligence Act 181 AI Act, art 29(4). 182 AI Act, arts 61(1), 62(1). 183 GDPR, 679, arts 77-78; Law Enforcement Directive, arts 52-53. 184 van Leeuwen (n 103) 20-21. 185 European Data Protection Board and European Data Protection Supervi- sor (n 47) para 18 (noting 'the absence of any reference in the text to the individual affected by the AI system appears as a blind spot in the Propo- sal' and the absence of any 'rights and remedies'). 186 GDPR, art 80. 187 Blue Guide, 104. 188 Market Surveillance Regulation, arts 11(3)(e), 11(7(a) (stating MSAs have only loose obligations to establish 'procedures for following up on complaints or reports on issues relating to risks or non-compliance' and to take into account complaints when taking a 'risk-based approach' to decide which checks to undertake). 189 .g. Case C-362/14
Cited by
The EU Artificial Intelligence Act (2024): Implications for healthcare
van Kolfschooten, H., & van Oirschot, J. (2024) · Health Policy · doi:10.1016/j.healthpol.2024.105152 · Cited by 314
Ebers, M., Hoch, V. R. S., Rosenkranz, F., Ruschemeier, H., & Steinrötter, B. (2021) · J · doi:10.3390/j4040043 · Cited by 153
Mission AI: The New System Technology
Sheikh, H. I., Prins, J. E. J., & Schrijvers, E. (2023) · Research for policy · doi:10.1007/978-3-031-21448-6 · Cited by 246
Understanding accountability in algorithmic supply chains
Cobbe, J., Veale, M., & Singh, J. P. (2023) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3593013.3594073 · Cited by 251
Self-citationGenerative AI and deepfakes: a human rights approach to tackling harmful content
Romero Moreno, F. (2024) · International Review of Law, Computers & Technology · doi:10.1080/13600869.2024.2324540 · Cited by 117
Metadata
- Authors
- Veale, M., & Zuiderveen Borgesius, F.
- Publisher
- Center for Open Science
- Published
- 2021-07-06
- OpenAlex
- W3203392755
- Cited by
- 91
- References
- 18
- Citations by year
- 2026: 31 · 2025: 11 · 2024: 8 · 2023: 18 · 2022: 16 · 2021: 6
- Type
- posted content
- Language
- en
- Open access
- green
- doi.org
- Locations
- UCL Discovery (University College London) (submittedVersion)
- SocArXiv (OSF Preprints) (submittedVersion, cc-by-nd)
- SocArXiv (OSF Preprints) (submittedVersion)
- Topics
- Social Sciences › Law › Digitalization, Law, and Regulation
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Physical Sciences › Computer Science › Artificial Intelligence › Law, AI, and Intellectual Property
- Keywords
- EU AI Act, AI regulation, risk-based regulation, enforcement regime, harmonization, AI policy, European Commission
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Demystifying the Draft EU Artificial Intelligence Act. https://app.cochat.ai/papers/10.31235/osf.io/38p5f
- The paper
- Veale, M., & Zuiderveen Borgesius, F. (2021). Demystifying the Draft EU Artificial Intelligence Act. https://doi.org/10.31235/osf.io/38p5f
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.