Preventing discrimination in the automated targeting of job advertisements

On the background of the increasing amount of discriminatory challenges facing artificial intelligence applications, this paper examines the requirements that are needed to comply with European non-discrimination law to prevent discrimination in the automated online job advertising business. This paper explains under which circumstance the automated targeting of job advertisements can amount to direct or indirect discrimination. The paper concludes with technical recommendations to dismantle the dangers of automated job advertising. Various options like influencing the pre-processing of big data and altering the algorithmic models are evaluated. This paper also examines the possibilities of using techniques like data mining and machine learning to actively battle direct and indirect discrimination. The European non-discrimination directives 2000/43/EC, 2000/78/EC, and 2006/54/EC which prohibit direct and indirect discrimination in the field of employment on the grounds of race or ethnic origin, sex, sexual orientation, religious belief, age and disability are used as a legal framework.     

Legal perspective on possible fairness measures – A legal discussion using the example of hiring decisions

With the increasing use of AI in algorithmic decision making (e.g. based on neural networks), the question arises how bias can be excluded or mitigated. There are some promising approaches, but many of them are based on a ”fair” ground truth, others are based on a subjective goal to be reached, which leads to the usual problem of how to define and compute ”fairness”. The different functioning of algorithmic decision making in contrast to human decision making leads to a shift from a process-oriented to a result-oriented discrimination assessment. We argue that with such a shift society needs to determine which kind of fairness is the right one to choose for which certain scenario. To understand the implications of such a determination we explain the different kinds of fairness concepts that might be applicable for the specific application of hiring decisions, analyze their pros and cons with regard to the respective fairness interpretation and evaluate them from a legal perspective (based on EU law).     

Dignity or discrimination: what paves the road towards equal recognition of same-sex couples in the European Union?

Abstract

The article analyses possibilities for the Court of Justice of the EU to go beyond its current narrow approach towards same-sex couples’ rights within the EU non-discrimination law framework, considering a comparative treatment of dignity-based arguments. It critically reviews the CJEU’s current approach exclusively focusing on direct discrimination and the comparator paradigm. By doing so, the Court has tolerated a situation of de facto discrimination and limited advancement of same-sex rights. The question is then whether the situation could be overcome if the CJEU would follow other courts and develop reasoning based on dignity to underpin the EU non-discrimination analysis with substantive meaning. The article rejects this proposition. Dignity is not suitable because it is both too wide and to narrow to ensure certainty and substantive protection within EU non-discrimination law. While the concept of dignity protects a minimum standard and can provide a floor of rights, non-discrimination law fosters equality by imposing procedural standards and challenging measures that effect groups differently. The concepts should thus not be conflated. Instead, a consistent application of the concepts of direct and indirect discrimination seems more promising.     

Taking Language Seriously: An Analysis of Linguistic Reasoning and Its Implications in EU Law

This article discusses legal reasoning at the European Court of Justice (ECJ). The following questions are addressed. First, the authors look at the way linguistic arguments are used in ECJ case‐law. Second, they consider whether the requirements of legal certainty, and more specifically that of predictability, may be fulfilled by reference to linguistic arguments in a multilingual legal system. The theoretical starting‐point is that of open‐endedness of language: no means exists to definitely pin down the meaning of words. Defining the meaning of words in a legal context is necessarily a matter of choice involving evaluative considerations. Consequently, when the ECJ uses linguistic arguments to justify a decision, it is an active agent choosing the meaning of words in a specific case. Essentially, the authors argue that legal reasoning based on linguistic arguments is particularly problematic from the viewpoint of legal certainty and predictability. In this respect, the key importance of systemic and teleological argumentation is emphasised in assuring convincing, acceptable and transparent legal reasoning especially in the context of multilingual EU law.     

The Evolving Judicial Politics of European Integration: The European Court of Justice and national courts revisited

This article analyzes how the judicial politics sparked by the European Union's (EU) legal development have evolved over time. Existing studies have traced how lower national courts began cooperating with the European Court of Justice (ECJ) to apply EU law because this empowered them to challenge government policies and the decisions of their domestic judicial superiors. We argue that the institutional dynamics identified by this ‘judicial empowerment thesis’ proved self‐eroding over time, incentivizing domestic high courts to reassert control over national judicial hierarchies and to influence the development EU law in ways that were also encouraged by the ECJ. We support our argument by combining an analysis of a dataset of cases referred to the ECJ with comparative case study and interview evidence. We conclude that while these evolving judicial politics signal the institutional maturation of the EU legal order, they also risk weakening the decentralized enforcement of European law.     

Challenging EU constitutional law: The Italian Constitutional Court's new stance on direct effect and the preliminary reference procedure

This article seeks to examine the relationship between EU law and the Italian legal order in light of the recent Italian Constitutional Court (ICC)’s jurisprudence attempting to redefine EU core principles. When fundamental rights are at stake, three assumptions are challenged: the determination of direct effect shall be a prerogative of the ECJ; EU directly effective provisions entail the disapplication of conflicting national law; judges have the discretion to refer preliminary references to the ECJ where a clarification on EU law is needed. The contribution argues that the judicial search for a balance between sovereignty and supranationality is undermined by the ICC's new resistance to the well‐established EU jurisprudence. In that respect, the paper posits that the ICC's activism is the result of an unjustified ‘argumentative self‐restraint’ of the ECJ vis‐à‐vis the evolution of EU foundational principles.     

Using statistics in employment discrimination cases.

Statistics play an important role in employment discrimination cases, and this role will expand as the legal profession becomes increasingly aware of the utility of the increasingly sophisticated statistical methods available. Historically, plaintiffs in employment discrimination cases have used statistics to establish prima facie cases concerning inequities in areas such as wage rates, personnel selection, promotion, layoff, and termination decisions. Defendants have also employed statistics to demonstrate the fairness of employment practices and policies--for example, by providing statistical evidence of the validity of a test used in personnel selection. This article provides an overview of the role of statistics and the major statistical techniques employed in discrimination cases.     

Principle of Proportionality as Principle of Economic Efficiency

The principle of proportionality is at the cornerstone of EU law, and precisely of the case‐law of the European Court of Justice (ECJ). In the law and economics literature, the general principles of law are commonly opposed to legal rules in terms of efficiency. On the one hand, the legal formalistic approach consists of apprehending the law as principled, whereby principles of law do not and should not encompass an efficiency rationale and should be self‐sufficient. On the other hand, the legal nihilism denying the existence or relevance of the general principles of law favours legal rules that are said to incorporate an efficiency rationale. I intend to analyse the efficiency rationale of probably the most important general principles of EU law—the proportionality principle. In this paper, I shall assert that not only does the EU proportionality principle encapsulate an efficiency rationale, but most importantly, it has been interpreted by the ECJ as such—hence, I propose the representation of the principle of proportionality as a principle of economic efficiency. After having introduced the principle of proportionality (1), I shall decipher the proportionality principle both from a law and economics perspective, and from a comparative perspective (2). Then, I shall delve into the jurisprudence of the ECJ so that the judicial reasoning of the Court as this reasoning proves the relevance of the proposed representation (3). Finally, I conclude in light of the findings of this paper (4).     

论人权公约中的禁止歧视

禁止歧视是联合国人权公约的重要内容,反对歧视的斗争也是联合国人权活动的中心,人权法与人权理论必须认真对待歧视的问题。歧视的核心规定性是“不合理的和主观的区分”,在人权公约中,禁止歧视有双重属性,既是附属性的,又是独立的。禁止歧视的理论基础是平等,核心是获得法律平等保护的权利。平等权的法律保障有积极和消极两个方面,国家负有不歧视的义务并有义务采取积极行动使人们免于歧视。     

On defense of “ethification” of law: How ethics may improve compliance with the EU digital laws

In recent years, academics and professionals witness the rise of the “ethification” of law, specifically in the area of ICT law. Ethification shall be understood as a proliferation of moral principles and moral values in the legal discourse within the areas of research, innovation governance, or directly enforceable rules in the industry. Although the ethical considerations may seem distant from mere regulatory compliance, the opposite is true. The article focuses on the positive side of the “ethification” of digital laws through the lens of legal requirements for impact assessments pursuant to General Data Protection Regulation and conformity assessments in the proposal for the Artificial Intelligence Act. Authors argue that ethical considerations are often absent in the context of using new technologies including artificial intelligence, yet they may provide additional value for organizations and society as a whole. Additionally, carrying out ethics-based assessments is already in line with existing regulatory requirements in the fields of data protection law and proposed EU AI regulation. These arguments are reflected in the context of facial recognition technology, where both data protection impact assessment under the EU General Data Protection Regulation and conformity assessment under the proposal of the EU Artificial Intelligence Act will be mandatory. Facial recognition technology is analyzed through the ethics-based assessment involving stakeholder analysis, data flows map, and identification of risks and respective countermeasures to show additional insights that ethics provides beyond regulatory requirements.     

The Function of the Proportionality Principle in EU Law

In this article the author assesses the proportionality principle in EU law from a legal theoretical and constitutional perspective with the aim of discovering the function of the principle. Having first discussed the implications of the proportionality principle being a general principle of law, and what function it has—namely to secure legitimacy for judicial decisions—the author suggests that there are several ways in which the principle can be interpreted. There is, nevertheless, a limit to this interpretation determined by the proposed function of the principle. In the third part of the article, the European Court of Justice's (ECJ's) interpretation of the principle is assessed. The assessment clearly shows that the ECJ is interpreting the principle in different distinguishable ways. The question could, however, be raised as to whether the ECJ in some areas is interpreting the principle in a way that undermines the very function of it.     

Levels of Generality in the Legal Reasoning of the European Court of Justice

The level of generality or of abstraction used to describe a precedent, a right, or the legislative intent behind a statutory provision or constituent purpose behind a constitutional provision can have a decisive impact on the outcome of a case. Characterising it in narrow terms has the effect of reducing the scope of decision of a judgment; conversely, a broader characterisation provides more leeway for a judge in a case to encompass its facts within the precedent, right or purpose in issue. The issue raised by the level of generality problem is the extent to which courts have a discretion or freedom of manoeuvre as to the level of generality they decide upon, and thus whether generality and abstraction are manipulable in the hands of judges and are not really predetermined by the legal sources in question or an established judicial method of interpretation. Uncontrolled judicial discretion of this kind is problematic from the point of view of the rule of law and democracy, especially when adjudication concerns constitutional provisions, the equivalent in the EU being interpretation by the European Court of Justice (ECJ) of the EU Treaties; reversal of ECJ interpretation through Treaty amendment is particularly difficult to achieve because it requires unanimous coordination by the Member States. This article examines two alternative ways of determining the correct or appropriate level of generality issue in ECJ interpetation, coherence or the legal traditions of the Member States, and argues in favour of the latter as a less subjective method. Application of the two alternative approaches is tested in two areas of EU law, state liability and criminal law.     

Supranational Fundamental Rights or Primacy of Sovereignty?

This analysis explores in detail various aspects of the possible legal impact of ‘British’ Protocol No 30 (the so‐called opt‐out from the EU Charter of Fundamental Rights). On the basis of a legal appraisal, it concludes that the Protocol is not in any way to be understood as a substantial derogation from the standard of protection of fundamental rights in the EU or as an ‘opt‐out’ from the Charter in a substantial sense. Nevertheless, its significance is definitely not to be underestimated. Its adoption as a source of primary law enshrines a legally binding interpretation of the Charter and, in particular, an interpretation of its horizontal provisions. In Article 1(2) and Article 2, the Protocol in fact confirms that the application of the Charter cannot lead to a change in the existing competencies framework. These provisions are of a declaratory nature and apply to all Member States. In Article 1(1), the Protocol is of a constitutive nature since it rules out an extensive interpretation of what can be considered national legal acts adopted in the implementation of EU law only for those States signed up to the Protocol. This specifically means that if, in the future, as part of the application of the Charter, the Court of Justice of the EU (ECJ) has a tendency to subsume a certain area of national legislation under the ‘implementation of Union law’ outside the field of implementing standards, in the spirit of the Ellinki Radiophonia Tileorassi judgment (and subsequently allow their reviewability with respect to their conformity with the Charter), such action would be admissible only for those Member States that have not acceded to the Protocol. However, the Protocol cannot exclude the continued application of the general principles of law instead of the positively constituted fundamental rights in the Charter by the ECJ.     

Reconsidering discrimination grounds in the data economy: An EU comparison of national constitutions

The principle of equal treatment (i.e., all people have the right to be treated equally) is protected by non-discrimination provisions in national constitutions across the EU as well as the EU Charter of Fundamental Rights (CFEU). These provisions specify which grounds (e.g., gender, race, religion) are prohibited to use as the basis for making decisions on people, such as offering a person a job. In the data economy, in which large amounts of personal data are collected and analyzed, it has become possible to make decisions on people on the basis of all kinds of grounds, also grounds that are not protected in anti-discrimination law (e.g., zip code, shoe size, wealth). Even though mostly unintentional, patterns revealed by sophisticated data analysis can turn out to be discriminatory, either directly or indirectly. Particularly indirect discrimination (i.e., discrimination by proxy) can be hard to discover and enforce. From a substantive perspective, these technological developments also raise the question which discrimination grounds should be protected, since discrimination grounds are in flux and not harmonized across the EU. In this paper, through legal comparison, discrimination grounds across EU national constitutions and the CFEU are compared, to identify overlaps and differences. This overview is then used to start the discussion on the extent to which current legislation is still appropriate in the data economy or should perhaps be reconsidered.     

The Lisbon Judgment of the German Constitutional Court,the Primacy of EU Law and the Problem of Kompetenz‐Kompetenz: A Conflict between Right and Right in Which There is No Praetor1

The ECJ has long asserted its Kompetenz‐Kompetenz (the question of who has the authority to decide where the borders of EU authority end) based on the Union treaties which have always defined its role as the final interpreter of EU law. Yet, no national constitutional court has accepted this position, and in its Lisbon Judgment of 2009 the German Constitutional Court (FCC) has asserted its own jurisdiction of the final resort' to review future EU treaty changes and transfers of powers to the EU on two grounds: (i) ultra vires review, and (ii) identity review. The FCC justifies its claim to constitutional review with reference to its role as guardian of the national constitution whose requirements will constrain the integration process as a standing proviso and limitation on all transfers of national power to the EU for as long as the EU has not acquired the indispensable core of sovereignty, i.e. autochthonous law‐making under its own sovereign powers and constitution, and instead continues to derive its own power from the Member States under the principle of conferral. Formally therefore, at least until such time, the problem of Kompetenz‐Kompetenz affords of no solution. It can only be ‘managed’, which requires the mutual forbearance of both the ECJ and FCC which both claim the ultimate jurisdiction to decide the limits of the EU's powers—a prerogative which, if asserted by both parties without political sensitivity, would inevitably result in a constitutional crisis. The fact that no such crisis has occurred, illustrates the astute political acumen of both the FCC and the ECJ.     

德国人工智能创造物的知识产权法保护及其启示

人工智能时代的到来,有关其创造物的知识产权法保护问题给现有的法律理念与规则带来了巨大挑战。因而比较法视阈欧盟背景下的德国经验或许有一定参考意义;即先通过评介欧盟层面与人工智能相关的法政策与伦理准则,再从德国层面整理其创造物知识产权法保护上的实务经验与理论讨论。进而从中获得对我国的启示:一方面,呼吁今后在我国人工智能创造物知识产权法保护的外部,进一步细化相关的法政策,并着手制定相适应的伦理准则;另一方面,在其内部,基于当前现行法修改实际,分别从专利法、著作权法二个主要维度提出具体的完善建议。并强调内外规制体系间的协调配合,以期塑造既具体可行又灵活完整的我国人工智能创造物的知识产权法保护体系。     

Directly Discriminatory Algorithms

Discriminatory bias in algorithmic systems is widely documented. How should the law respond? A broad consensus suggests approaching the issue principally through the lens of indirect discrimination, focusing on algorithmic systems’ impact. In this article, we set out to challenge this analysis, arguing that while indirect discrimination law has an important role to play, a narrow focus on this regime in the context of machine learning algorithms is both normatively undesirable and legally flawed. We illustrate how certain forms of algorithmic bias in frequently deployed algorithms might constitute direct discrimination, and explore the ramifications—both in practical terms, and the broader challenges automated decision-making systems pose to the conceptual apparatus of anti-discrimination law.     

The European Union and Penal Law

Technically-speaking, penal law remains outside the competence of the European Communities and Union. However, mirroring other legal developments within Europe, a combination of higher Community 'principles' such as proportionality, non-discrimination, free competition and loyal co-operation, together with secondary Community law, has on the one hand, led to an unforeseen process of the harmonisation of national penal systems; with national norms either being set aside by Community law, or given extended scope in the pursuance of EC/EU goals. On the other hand, certain European interests – most notably, the need to safeguard the European Union budget – have proven strong enough to prompt the evolution of a nascent penal law of the EU; the most noteworthy development here being the drawing up of an independent European 'corpus juris' covering penal policy and procedure in the area of EU budget protection.     

Why Do the European Court of Justice Judges Need Legal Concepts?

The article considers the reasons why the European Court of Justice (ECJ) judges need legal concepts when they pronounce their judgments. It points out that the ECJ as a law‐interpreting and an ipso facto law‐making court needs legal concepts to communicate results of its interpretative and law‐making enterprise. The article also shows how in the context of Article 234 EC preliminary ruling procedure legal concepts become useful tools of portraying ECJ judgments as mere products of interpretation and not as the results of subsuming the facts of the case into a legal provision. It is by means of application of legal concepts, that the ECJ judges are able to justify that they are not overstepping the mandate they have been entrusted with. In the same time the use of legal concepts enables them to engage in dialogue with national judges, who seek guidance as to the content of EC law rules, and to maintain a strong doctrine of precedent. Most importantly, however, the use of concepts promotes coherence which, the article maintains, is the primary source of Community law's authority, and thus constitutes the foundational technique of persuading the relevant audience that Community law is indeed a legal system.