2023 | Buch

Governmental Automated Decision-Making and Human Rights

Reconciling Law and Intelligent Systems

verfasst von: Stefan Schäferling

Buchreihe : Law, Governance and Technology Series

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Inhaltsverzeichnis

Über dieses Buch

Mit den wachsenden Fähigkeiten der künstlichen Intelligenz integrieren Regierungen KI-Technologien in die Entscheidungsfindung in der Verwaltung und sogar in der Justiz, unterstützen menschliche Entscheidungsträger und ersetzen sie in einigen Fällen sogar. Predictive Policing, automatisierte Verwaltung der Sozialleistungen und automatisierte Risikobewertung bei der Strafverfolgung sind nur einige prominente Beispiele für einen allgemeinen Trend. Obwohl die Hinwendung zu staatlich automatisierter Entscheidungsfindung verspricht, die Auswirkungen menschlicher Vorurteile zu verringern und Effizienzgewinne zu erzielen, birgt die Verringerung des menschlichen Elements in der staatlichen Entscheidungsfindung auch erhebliche Risiken. Dieses Buch analysiert diese Risiken durch eine vergleichende Verfassungs- und Menschenrechtslinse und untersucht das US-Recht, das deutsche Recht und das internationale Menschenrechtsrecht. Sie unterstreicht auch die strukturellen Herausforderungen, die die Automatisierung für Rechtssysteme darstellt, die auf der Annahme ausschließlich menschlicher Entscheidungsfindung beruhen. Besondere Aufmerksamkeit wird der Frage gewidmet, ob bestehende Gesetze den Mangel an Transparenz bei der automatisierten Entscheidungsfindung durch Regierungen, ihre diskriminierenden Prozesse und Ergebnisse sowie ihre grundlegende Herausforderung an menschliches Handeln adäquat beheben können. Aufbauend auf dieser Analyse schlägt er einen Weg zur Sicherung der Werte der Menschenwürde und des Handelns im Herzen demokratischer Gesellschaften und der Rechtsstaatlichkeit in einer zunehmend automatisierten Welt vor. Dieses Buch wird für Forscher und Wissenschaftler interessant sein, die sich mit der sich entwickelnden Beziehung von Recht und Technologie sowie Menschenrechtsaktivisten befassen. Darüber hinaus stellt sie einen wertvollen Beitrag zur Debatte über die Regulierung der künstlichen Intelligenz und die Rolle dar, die die Menschenrechte in diesem Prozess spielen können.

Mit KI übersetzt

Inhaltsverzeichnis

Frontmatter

Approaching Governmental Automated Decision-Making and Human Rights

Frontmatter

Chapter 1. Introduction

Abstract

Based on the example of automated risk assessments in criminal sentencing, this chapter illustrates the practice of governmental automated decision-making and the challenges it raises from a regulatory perspective. Considering the speed with which technology is advancing, it is important to identify the core normative values at risk, which can serve as the basis for a more detailed regulatory framework in the face of increased automatization of governmental decisions. The chapter introduces this book’s proposal for a path towards a right protecting human agency and personhood against the threat of governmental automated decision-making.

Stefan Schäferling

Chapter 2. Artificial Intelligence and Automated Decision-Making: Terminological Clarifications

Abstract

The growing popularity of automated decision systems is based on recent advances in an area of computer science that is described by the enigmatic term artificial intelligence (AI). While the technology has come a long way since its origins in the 1950s, little progress has been made on terminological clarity. This chapter traces the development of the technology and decribes the basic processes behind AI. It explains why AI is not a helpful term from a regulatory perspective and suggests placing the problematic act of decision-making at the center of the further analysis.

Stefan Schäferling

Chapter 3. Automated Decision-Making and the Law

Abstract

Governing modern rule-based societies involves making countless legal decisions. Thus far, these decisions have been made exclusively by humans. However, the human decision-making process does not always correspond to the ideal image of fact based objective legal reasoning. This chapter introduces different theories on human decision-making and points out its strengths and flaws. It also shows how the functioning of human decision-making processes as the default for legal decisions have influenced and shaped the law. Automated decision-making is introduced as a potential alternative, describing its functioning and its limits. This chapter also examines some of the reasons for the rise of automated decision-making despite its weaknesses. Finally, it examines current and possible future applications of automated decision-making by the executive and judicial branches in a case study ranging from predictive policing to automated administration of benefits to automated decision-making technology’s use in courts.

Stefan Schäferling

Part II

Frontmatter

Chapter 4. Identifying Challenges of Governmental Automated Decision-Making

Abstract

Building on the previous case studies this chapter identifies risks associated with the increasing automation of government decision-making. On a factual level, human’s automation bias plays an important role. On the legal level, two challenges stand out: discrimination and intransparency. The chapter explains how bias can enter the allegedly objective automated decision-making process and explains the different ways automated decision-making exacerbates the opacity of governmental decision-making.

Stefan Schäferling

Chapter 5. Confronting the Legal Challenges of Governmental Automated Decision-Making

Abstract

This chapter undertakes an in-depth comparative legal analysis, examining the ability of US and German (constitutional) law as well as of international human rights law to address the challenges of discrimination by and opacity of governmental automated decision-making. It identifies strengths and weaknesses of the different legal systems’ approaches and offers suggestions for possible reforms to counter the difficulties existing law faces in providing protection against discriminatory and opaque outcomes of automated decision-making processes.

Stefan Schäferling

Chapter 6. The Underlying Challenge to Human Agency

Abstract

Despite their prevalence, the problems of algorithmic discrimination and opacity are ultimately only symptoms of a deeper concern about automated decisions as such. After all, humans also do not always act with full transparence and show a propensity to discriminate. Over the centuries, however, the law has developed a sophisticated framework that provides a level of transparency and protection against discrimination that, while not complete, appreciably reduces the risk of violations of individuals’ rights. This framework, however, rests on the assumption of human actors and human agency. For automated decisions, there remains a fundamental legal disconnect that transparency obligations and anti-discrimination laws cannot fully address. Thus, it is not their propensity to discriminate in new ways or their lack of transparency that distinguishes automated decisions, but rather their fundamental lack of humanity. This chapter addresses this fundamental challenge to human personhood and agency.

Stefan Schäferling

Part III

Frontmatter

Chapter 7. The Case for a Right Against Automated Decision-Making

Abstract

Building on the diagnosed unease and legal concerns with governmental automated decision-making, this chapter turns to the question of whether a general right against automated decisions to protect human personhood and agency can serve as a starting point for further regulation. It demonstrates the insufficiency of other regulatory tools before showing that in some legal systems, dignitary concerns provide a sound legal basis for such a proposed right.

Stefan Schäferling

Chapter 8. Reconciling Human Rights and Intelligent Systems: Contents and Implications of a Right Against Governmental Automated Decision-Making

Abstract

Having set out a dogmatic basis for a right against governmental automated decision-making in the European legal context, this chapter elaborates on such a right’s content and role in a future legal framework. An undifferentiated approach that categorically excludes automation will not do justice to the multifaceted and complex issues involved, especially given the weaknesses of human decision-making. Instead, it is necessary to balance the conflicting interests and attempt to reap the benefits of automated decision making without at the same time unduly compromising human autonomy. A proposal for a right is made that embraces different spheres of risk, allowing the necessary precautions to be adapted to the individual case.

Stefan Schäferling

Chapter 9. Conclusion and Outlook

Abstract

This chapter concludes by summarising the findings and calls for not carelessly sacrificing hard-won human rights achievements for efficiency gains, despite all the advantages that technological developments bring with them, recalling that, ultimately, law is meant to serve the good of humankind.

Stefan Schäferling

Titel: Governmental Automated Decision-Making and Human Rights
verfasst von: Stefan Schäferling
Verlag: Springer Nature Switzerland
Electronic ISBN: 978-3-031-48125-3
Print ISBN: 978-3-031-48124-6
DOI: https://doi.org/10.1007/978-3-031-48125-3