DTA

Digital Theses Archive

 

Tesi etd-03142022-140108

Type of thesis
Dottorato
Author
BIBER, SUMEYYE ELIF
URN
etd-03142022-140108
Title
A Rights-Based Inter-Legal Approach to the Human and Fundamental Rights Challenges posed by Artificial Intelligence Systems
Scientific disciplinary sector
IUS/20
Course
Istituto di Diritto, Politica e Sviluppo - PHD IN LAW
Committee
relatore Prof. PALOMBELLA, GIANLUIGI
Keywords
  • AI-Act
  • artificial intelligence systems
  • fundamental rights
  • human rights
  • inter-legality
Exam session start date
24/11/2022;
Availability
parziale
Abstract
The thesis examines how European Union and international law and its regulatory models address the fundamental and human rights challenges posed by artificial intelligence (AI) systems. In particular, the 2021 proposed Artificial Intelligence Act, General Data Protection Regulation, and EU fundamental rights and international and European human rights law are considered and assessed to the extent that can tackle such challenges. Illustrative case studies of AI systems - concentrating on the most significant AI practices affecting the protection of fundamental and human rights - are chosen to demonstrate these challenges. It is argued that the current world’s first approach of the EU focusing on a risk-based regulation is not sufficient to protect fundamental and human rights. Given the core purpose of the EU that is to secure fundamental values and principles throughout the production process of AI, from the design phase to its deployment, the theory of inter-legality and the methodology it presents are proposed for a new approach called “a rights-based inter-legal approach”. The thesis provides five solid reasons, why this approach is highly relevant to the issue of AI. Further, it shows how it can be applied in three steps both at the legislation and adjudication levels.<br>First, the rapidly increasing proliferation of extra-state regimes and international organizations focusing on AI proves that AI is a global issue. At the international level, there is a new initiative called “Global.AI Policy” to help policymakers navigate the international AI governance framework and access the necessary tools and best practices. It is an online platform targeting cooperation between intergovernmental organizations. As of September 2021, the platform has eight partners namely, the Council of Europe, the European Commission, the European Agency for Fundamental Rights, the Inter-American Development Bank, OECD and the OECD.AI, the UN, UNESCO, and the World Bank Group (WBG). <br>Second, an inter-legal rights-based approach is necessary due to the extremely dynamic characteristics of AI. Closed legal systems are not able to capture such a dynamic field and the changes it causes in society. Such dynamism requires up-to-date approaches. In this sense, an inter-legal approach is promising as it excludes the monopoly or “legal hegemony” of a single legal order. <br>Third, proposing AI as a solution for other great challenges makes AI materially interconnected to these other global concerns. It would be naïve to not consider global security while focusing on the regulation of AI-driven surveillance technologies such as biometric identification or categorization systems. Axelrod and Kehoane identified this connection as “issue-linkage”: “Most issues are linked to other issues. This means that games being played on different issues-different “chessboards,” in Stanley Hoffmann&#39;s phrase affect one another. Connections between games become important when issues are linked”. <br>Fourth, AI presents highly complex challenges. The main claim of legal scholars is that the law remains inadequate to cope with such challenges. Closed legal systems would not be able to solve the problems that AI entails. It is predictable that such a wide and inter-connected issue will inevitably test the limits and the scope of legal regimes. However, inter-legality requires questioning the law of the case that should be searched within the interwoven legal texture.<br>Fifth, the inadequacy of law in the face of artificial intelligence might create a sort of “lawless” area. Inter-legality is able to cope with this “gap”. Such an approach is promising as it offers a bridge between legal systems by revealing the connections between those systems. From an inter-legal perspective, the diversity of jurisgenerative entities, regimes, or orders all contribute to bridging the gap, making the composite law governing the issue at stake. <br>
Files