To date, all stakeholders are working intensively on policy design for artificial intelligence. All initiatives center around the requirement that AI algorithms should be fair. But what exactly does it mean? And how can algorithmic fairness be translated to legal and policy terms? These are the main questions that this paper aims to explore. Each discipline approaches those questions differently. While computer scientists may favor one notion of fairness over others across the board, this paper argues in favor of a case-by-case analysis and application of the relevant fairness notion. The paper discusses the legal limitations of the computer science (CS) notions of fairness and suggests a typology of matching each CS notions to its corresponding legal mechanism. The paper concludes that fairness is contextual. The fact that each notion, or group of notions, correspond with a different legal mechanism, makes them suitable for a certain policy domain more than others. Thus, throughout the paper, examples for possible applicability of the CS notions to some policy domains will be introduced. In addition, the paper will highlight for both developers and policymakers the practical steps that need to be taken in order to better address algorithmic fairness.

In some instances, notions of fairness that seem, on their face, unproductive from a technical perspective, could in fact be quite helpful from a legal perspective. In other instances, desirable notions in the eye of computer scientists could be challenging to implement in the legal regime, due to the need to determine complex moral and legal questions. Thus, as the article emphasizes, a one-size-fits-all solution is not applicable for algorithmic fairness. Rather, an approach that demonstrates a deep understanding of the specific context that a certain algorithm is operating in can guarantee a fairer outcome.

Despite the recognized importance of witness testimony in addressing systematic violence and human rights violations, reflected in the participation of large numbers of witnesses in international legal processes, establishing facts based on oral testimonies in international criminal tribunals remains a contentious matter.

The article develops a new model for assessing judicial attention to and engagement with testimonial narratives, in particular of victims of sexual violence, by conceptualizing the testimonies as “textual datasets.” This article presents the results of an algorithm-based approach for analyzing testimonial corpora, applying a generative statistical model known as unsupervised topic modeling. I employ LDA topic modeling for empirically assessing the international courts’ capacity to “listen” to large quantities of witness testimonies. Harnessing the large number of testimonies in international criminal trials, I use topic modeling in order to explore latent themes and semantic fields that could benefit the legal process and its critical scholarly appreciation.
This article proposes Automated Content Analysis, in particular topic modeling method, as a novel method to assist scholars and practitioners in making sense of complex legal cases, involving large amounts of testimonies, documents, and data, while preserving the voice and vocabulary of the individual witness.

This article proposes Automated Content Analysis, in particular topic modeling method, as a novel method to assist scholars and practitioners in making sense of complex legal cases, involving large amounts of testimonies, documents, and data, while preserving the voice and vocabulary of the individual witness.

This article highlights the potential of topic modeling methods, rooted in Natural Language Processing and Digital Humanities, to overcome critical impediments in empirical legal studies. It demonstrates the method’s capacity to transform both as a practical heuristic mechanism that can be employed during the legal proceeding, and in its ex-post analysis in legal scholarship.

While the sharing economy brings significant social benefits in China, it comes with regulatory challenges that are novel and unpredictable. How should regulators handle these challenges? This paper offers fresh insights into the regulatory approach to the ride-hailing industry, the most comprehensively regulated sharing industry in China. A historical review identifies three regulatory approaches deployed to date: self-regulation, market-based regulation and government regulation. Self-regulation relies on the platforms with incentive to provide better service for greater profit, and to deal with sharing-specific challenges. Market-based regulation invites rivals to keep a watchful eye on other players, in order to enhance their market position by outperforming the competition. Both approaches are capable of delivering quick, and often innovative, responses to new challenges. Government regulation, on the other hand, came late and plays a neutral role. The rules there are mostly of the “old wine in a new bottle” kind; in other words, applying existing (old) rules to the new sharing economy. Those rules could contribute to a level playing field for traditional and sharing-market players if managed properly. This article argues that government regulations are inadequate for solving sharing-specific challenges such as the legal status of the participants, the challenges of uncertain externalities, and new forms of competition. Instead, regulators should in the future give more affirmative value to self-regulation andmarket-based regulation. These complementary approaches are capable of yielding innovative and sharing-specific regulatory responses, from which the government regulators can glean and evaluate before codifying them.

This article examines third-party liability issues of automated vehicles (AV) as currently regulated in Europe at the level of the European Union (EU) and at the national level.

We begin with a brief presentation of international law on traffic rules, whose binding effect has influenced the content of European provisions. We proceed with the analysis of the provisions of the Product Liability Directive and the Motor Insurance Directive in view of their applicability to AV. Subsequently, we briefly analyze the pertinent provisions of German and English law on product liability and road traffic liability, including the special rules of the German Road Traffic Act, which focuses on the liability of the keeper and the driver, as well as the English Automated and Electric Vehicles Act 2018, which focuses on insurance and represents a different approach. We then outline the legal landscape in the US and compare it with the European. Afterwards, we briefly examine less obvious parameters, such as human-factors, the role of media and ethics, and explore the potential for international harmonization.

We conclude that, at present, the risks and benefits from the use of AV are not making a convincing case to depart from traditional liability rules on road traffic and defective products. There is no uninsurable disaster potential and no radical change in people’s lives to justify limiting the legal right of uninvolved victims to receive compensation compared to ordinary vehicles. There are more appropriate means than liability reform to incentivize technological development. Moreover, establishing uniform international liability rules would be desirable, but appears neither necessary nor politically feasible.