Martin Ebers and Susana Navas (eds) Algorithms and Law reviewed by Francisco de Elizalde

• DOI: https://doi.org/10.13169/prometheus.37.4.0385
• Pages: 385-386
• Published date: 28 March 2022
• Date: 28 March 2022
• Author: Francisco de Elizalde

Book review385

BOOK REVIEW

Algorithms and Law, edited by Martin Ebers and Susana Navas (2020) Cambridge University

Press, Cambridge, 297pp., hardback £85, ISBN: 9781108424820.

Law sails as tankers do: steadily and with difficulty in changing direction. The relative inflexibility

of law enhances legal certainty as an essential requirement for the protection of rights and, eventu-

ally, the evolution of societies. Legal thought and legal analysis usually march at the same prudent

pace. However, the technological disruption that artificial intelligence (AI) brings has created

turmoil. The ‘law tanker’ suddenly started to behave as a dinghy in a hurricane. An institution as

accustomed to serious, long and technical legal drafting as the European Parliament introduced a

legal instrument with references to Mary Shelley’s Frankenstein, Pygmalion and the story of

Prague’s Golem (see European Parliament, 2017). The legal debate tried to respond to unusual

questions on the feasibility of an ‘electronic personality’ for robots to address liability – whether the

phenomenon should be tackled with ethical (i.e., not legally enforceable) rules instead of regulation

or even if an entire new legal framework was needed for AI.

Far away from these positions, some legal scholars rapidly framed the challenges that AI

and new technologies pose to law. The editors of this book were among them. In a relatively short

period, traditional legal thinking has become mainstream again, pushing tabula rasa to the edge. In

a nutshell, current debate increasingly assumes that law is fit to tackle the challenges AI presents,

though with changes. The book under review is an important milestone in this direction.

The reader will find in this collection a legal assessment of self-learning algorithms, primarily

from the perspective of European Union (EU) law. In chapter 1, Sami Haddadin and Dennis Knobbe

provide an overview of AI from a technical (i.e., non-legal) angle. It is a good opening for readers

not familiar with the topic, one which underscores the realistic approach of the book. Chapter 2, by

Martin Ebers, together with chapters 1 and 3, could have formed a separate part of the book because

of their introductory character. They are rich in appealing questions. Chapter 2 first addresses those

characteristics of AI that pose common challenges to all areas of law: the move from causation to

correlation, the growing autonomy of AI systems and the problem black boxes pose for law, which

requires transparent and reasoned decisions. The chapter introduces ethical questions on the interac-

tion of algorithms and human beings and provides an overview of specific legislation that would

apply to AI. Chapter 3, by Mario Martini, focuses on the opacity of machine-learning systems,

discrimination and monopolization of market power and knowledge (for which the author proposes

legislative measures).

The following chapters (4–10) tackle specific legal issues arising from the use of self-

learning algorithms and robotics. In chapter 4, Diana Sancho focuses on automated decision-making,

which is regulated in Article 22 of the European General Data Protection Regulation (GDPR). The

author presents an expansive interpretation of the provision, interesting in view of the current tech-

nical reality whereby solely automated decisions (i.e., those literally covered by the provision) are

infrequent. Chapter 5, by Susana Navas, and chapter 6, by Ruth Janal, explore tortious liability for

autonomous systems, addressing the complexities that they introduce in determining the individuals

liable for wrongs if compared, for example, with a normal product. In chapter 7, Gerald Spindler

deals with algorithms in high-frequency trading against the backdrop of financial markets while

assessing whether this sector can serve as model for others. In chapter 8, Susana Navas delves into

copyright law. Taking an innovative stance, she offers convincing arguments for the protection of

works created by autonomous systems. Chapter 9, by Brian Subirana, Renwick Bivings and Sanjay

Sarma, introduces ‘wake neutrality’ of voice-recognition systems in open conversational commerce

DOI:10.13169/prometheus.37.4.0385