Managing Discrimination Risks in Artificial Intelligence under EU Law: A Legal Assessment of the Effectiveness of the AI Act

Hilal Aydın

TABLE OF CONTENTS

Chapter 1
INTRODUCTION
1.1. Background and Research Question
1.2. Thesis Objectives
1.3. Research Methodology
1.4. Scope and Limitations
1.5. Thesis Structure

Chapter 2
THE PRINCIPLE OF NON-DISCRIMINATION AND ITS LEGAL GROUNDS
2.1. The Legal Status of the Principle of Non-Discrimination in Public International Law and Its Normative Foundations
2.2. The Principle of Non-Discrimination in EU Law: Legal Status and Underlying Norms
2.2.1. Discrimination Grounds
2.2.2. Discrimination Types

Chapter 3
DISCRIMINATION ARISING FROM AI SYSTEMS UNDER EU NON-DISCRIMINATION LAW
3.1. Definition, Development, and Functioning of AI and Its Fields of Use
3.2. AI and the Risk of Discrimination
3.2.1. Algorithmic Bias and Its Underlying Mechanisms
3.2.2. Examples of Discriminatory Outcomes Caused by AI Systems
3.3. The Scope, Application and Limitations of EU Non-Discrimination Directives in the Context of AI
3.4. CJEU Discrimination Case Law in Relation to AI Systems
3.4.1. Determination of Direct Discrimination by the CJEU
3.4.2. Determination of Indirect Discrimination by the CJEU
3.4.3. Application of the Burden of Proof

Chapter 4
THE AI ACT AND ITS GOVERNANCE OF DISCRIMINATION RISKS IN AI SYSTEMS
4.1. The Emergence Process of the AI Act
4.2. An Overview of the AI Act
4.2.1. The Legal Status of the AI Act
4.2.2. The Purpose and Subject Matter of the AI Act
4.2.3. The Scope of the AI Act
4.3. The Risk-Based Approach of the AI Act: Effectiveness in Regulating Discrimination Risks
4.3.1. Prohibited AI Practices
4.3.2. High-Risk AI Systems
4.3.3. Tools for Preventing Discrimination in High-Risk AI Systems under the AI Act
4.3.4. Certain AI Systems and Their Transparency Requirements
4.3.5. Minimal-Risk AI Systems
4.4. General-Purpose AI Models

Chapter 5
CONCLUSION