Demystifying the Draft EU Artificial Intelligence Act — Analysing the good, the bad, and the unclear elements of the proposed approach

  title={Demystifying the Draft EU Artificial Intelligence Act — Analysing the good, the bad, and the unclear elements of the proposed approach},
  author={Michael Veale and Frederik J. Zuiderveen Borgesius},
  journal={Computer Law Review International},
  pages={97 - 112}
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… 

Tables from this paper

Artificial Intelligence ante portas: Reactions of Law
Artificial intelligence and algorithmic decision-making causes new (technological) challenges for the normative environment around the globe. Fundamental legal principles (such as non-discrimination,
Gradient Legal Personhood for AI Systems—Painting Continental Legal Shapes Made to Fit Analytical Molds
What I propose in the present article are some theoretical adjustments for a more coherent answer to the legal “status question” of artificial intelligence (AI) systems. I arrive at those by using
The Problem of Zombie Datasets: A Framework For Deprecating Datasets
This paper examines the public afterlives of several prominent deprecated or redacted datasets, including ImageNet, 80 Million Tiny Images, MS-Celeb-1M, Duke MTMC, Brainwash, and HRT Transgender, in order to inform a framework for more consistent, ethical, and accountable dataset deprecation.
There is an elephant in the room: Towards a critique on the use of fairness in biometrics
It is proved that biometric fairness criteria are mathematically mutually exclusive and illustrated empirically that a fair biometric system is not possible by reproducing experiments from previous works, and the politics of fairness in biometrics systems is discussed by situating the debate around biometric systems used at the border.


Demystifying the Draft EU Artificial Intelligence Act
An overview of the draft AI Act is presented and its implications are analyzed, drawing on scholarship ranging from the study of contemporary AI practices to the structure of EU product safety regimes over the last four decades.
The right to a human in the loop: Political constructions of computer automation and personhood
Two different legal constructions of automated data processing are examined, one that has furnished a right to a human in the loop that is intended to protect the dignity of the data subject and the other that promotes and foster full automation to establish and celebrate the fairness and objectivity of computers.
The Unfair Commercial Practices Directive
Abstract In an examination of the main provisions of Directive 2005/29, the solutions adopted for a number of the key design issues are assessed. First, has the Directive overcome the problem of the
Algorithms that remember: model inversion attacks and data protection law
  • Michael Veale, R. Binns, L. Edwards
  • Computer Science, Medicine
    Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences
  • 2018
Recent work from the information security literature around ‘model inversion’ and ‘membership inference’ attacks is presented, which indicates that the process of turning training data into machine-learned systems is not one way, and how this could lead some models to be legally classified as personal data.
Enslaving the Algorithm: From a “Right to an Explanation” to a “Right to Better Decisions”?
This work outlines recent debates on the limited provisions in European data protection law, and introduces and analyze newer explanation rights in French administrative law and the draft modernized Council of Europe Convention 108.
European integration through standardization: How judicial review is breaking down the club house of private standardization bodies
Standardization is slowly but slowly but surely marching forward into the public domain. The revision of the European standardization package is stimulating this development because the use of
Regulating Code - Good Governance and Better Regulation in the Information Age
  • I. Brown, C. Marsden
  • Political Science, Computer Science
    Information revolution and global politics
  • 2013
The authors describe the increasing "multistakeholderization" of Internet governance, in which user groups argue for representation in the closed business-government dialogue, seeking to bring in both rights-based and technologically expert perspectives.
Platform values and democratic elections: How can the law regulate digital disinformation?
The effects that disinformation initiatives have on freedom of expression, media pluralism and the exercise of democracy are examined, from the wider lens of tackling illegal content online and concerns to request proactive (automated) measures of online intermediaries.
Case C-171/11 SpA v Deutsche Vereinigung des Gas- und Wasserfaches
The issue of the horizontal direct effect of the free movement provisions has the tendency to cause anxiety among lawyers who fear for the fate of private autonomy. The issue also evidently troubles
Consent Management Platforms under the GDPR: processors and/or controllers?
It is concluded that CMPs process personal data, and two major CMP providers in the EU: Quantcast and OneTrust are paired with a legal analysis, and multiple scenarios wherein C MPs are controllers are identified.