Book Review: Robot Rules, Regulating Artificial Intelligence

Artificial Intelligence is a popular topic that is building momentum, whether you are already interested in AI and looking for more discussion or if you want to join the conversation, this book Robot Rules, Regulating Artificial Intelligence by Jacob Turner is great place to start. Jacob is a Barrister at Fountain Court, specialising in AI among other things.

Until now, the introduction of the book says, laws have only had one subject: humans. The rise of artificial intelligence (AI) presents novel issues for which current legal systems are only partially equipped. Who should be liable for harm caused by AI? Is it wrong to destroy a robot? Can AI be made to follow moral rules?

This book provides a roadmap for these and other related question, considering what new regulation is required, including looking at what the rules should be, who should shape them and how they should be upheld.

The book is presented in 8 chapters. Chapter 1 provides an introduction to AI. First it uncovers the origins of modern AI research in 1956 and from wider sources such as science fiction [“I’m sorry Dave, I cannot do that” HAL]. The it turns to defining AI. Turner explains that AI can be distinguished by two classifications: narrow and general. Narrow (or weak) AI is where a system is able to achieve certain specified goals such as translation or navigation, whereas general (or strong) AI has the ability to achieve an unlimited number of goals and even independently set new goals. The chapter then focuses on the meaning of “intelligence” in the context of AI. Turner highlights the importance of a definition in order to provide for effective legal regulation that is specific and workable. As such, proposes the following definition AI is:

the ability of a non-natural entity to make choices by an evaluative process.

From there, Turner suggests that there are three main areas in which AI will give rise to new challenges which are addressed in the subsequent chapters:

1. Responsibility: If AI were to cause harm, or to create something beneficial, who should be held responsible?
2. Rights: Are there moral or pragmatic grounds for granting AI legal protections and responsibilities?
3. Ethics: How should AI make important choices, and are there any decisions it should not be allowed to take?

Chapter 2 focuses more specifically on the unique factors of AI, elaborating on AI as a legal phenomenon. This calls into question certain fundamental assumptions within legal systems, such as subjects and agency or causation. Turner argues that AI is unlike any previous technology because it is not fixed or static, and likens the maturity of AI technology to the point when children reach the age of reason and become legally responsible for their actions. 

This leads to chapter 3, which considers the responsibility for AI, in terms of liability for harm and of particular interest to readers, accountability to for creative outputs such as authorship. Turner discusses various mechanisms for establishing who or what is responsible when AI causes harm or creates benefit, relating to negligence, contracts, criminal law, insurance and copyright. 

Hello, HAL. Do you read me, HAL?!
Image: Michael Berch

Turner argues that AI is capable of original creative expression through independent choice, sequence and combination and therefore able to create copyright works. However, the law requires that the owner of the copyright be a legal person, except in the instance of computer-generated works where the copyright holder is the person who made the necessary arrangements for the creation to be undertaken [s9(3) CDPA 1988]. 

Turner poses the question of whether we should grant copyright to AI and this is address in Chapter 4. The chapter focuses on moral reason for granting rights to AI. It builds this argument by looking at the development of animal rights and human rights. Taking this argument further, chapter 5 discusses the arguments for and against granting AI legal personality; in the sense of a technical label for a bundle of rights and responsibilities.

Chapter 6 proposes ways to design international systems to create the types of new laws needed to respond to AI developments. It proposes how we ought to design, implement and enforce new rules tailored to AI. In particular, Turner suggests that these rules cannot be created on a national level and instead there should be an agreed global code. 

A Kat in Space Odyssey
Image: Mumes World

The last two chapters of the book focus on ways to prevent undesirable consequences. Chapter 7 considers the regulation of the human creators of AI, in particular the legal and ethical standards of those who create, design, programme, operate and collaborate with AI. Chapter 8 considers the possibilities of building in or teaching AI rules for self-regulation. 

In the forward, The Rt Hon Lord Neuberger describes it as a thoughtful and informed book which analysis the implications of current and future developments in AI and how we should plan to deal with them. AI is a popular topic of discussion – some might even say buzzword – but this is the first book this Kat has read that takes a pragmatic view on how regulation should and can be made. The book offers valuable suggestions and practical solutions. As Turner puts it: 

Today’s regulation is likely to influence how technology develops. In building structures for effective everyday legal regulation in the medium term, we can prepare ourselves far better…

Therefore, this book would appeal to anyone with an interest in AI and the future of the legal system in responding to these types of technological developments. The book is presented in a readable and informative manner that would be accessible to experts and non-experts alike.

ebook ISBN 978-3-319-96235-1  - €23.79
softcover ISBN 978-3-319-96234-4 - €29.11
Available from Palgrave Macmillan 
Available from Amazon