Darren Grayson Chng reviews a new book looking into the of legal and ethical issues arising from developments in AI and robotics.
Artificial Intelligence, Robots and the Law’ is a book written by two Australians on legal and ethical issues arising from developments in AI and robotics. It is described as a call to students, lawyers, technologists, academics, regulators, and others to engage with the issues raised by new developments in these technologies. Except for the book’s first and last chapters which provide an introduction and conclusion, you can think of the remaining 9 chapters as papers on different topics relating to AI and robotics.
After flipping a couple of pages, Chapter 1 brings the reader on “a brief tour of technical concepts”. This heading is aptly named, for the section provides more than just an explanation of what a concept means in the technical sense, but also gives the reader a bit of its history, and how it has been used in practice.
Chapter 2 is titled ‘AI, Robotics, and Norm Building’. It discusses examples of mischief in AI and robotics e.g., deliberate misuse and government surveillance, then talks about common ethical principles. The last section of the chapter is on ‘hardening ethical principles into legal norms’. This is not a topic often written about, and so I wish that the book spent more than two paragraphs on it.
Chapter 3 talks about technological change and regulatory effectiveness. It spends a good couple of pages on theoretical frameworks for regulating technology, and towards the end, raises the ‘pacing problem’, which refers to the slow pace of law’s reaction to new technology, resulting in a gap or void where no law applies to the technology. People who complain about the pacing problem usually sensationlise the issue, and I agree with the book’s view that these people gloss over the fact that technologies rarely operate in unregulated terrain. Many times there are existing sectoral laws or instruments that can apply to the technology, and depending on the sectoral regulator’s processes, the regulator may be able to amend their regulations or take a soft regulatory approach to manage the new technology, while the government evaluates it and the most appropriate way to respond to arising problems.
Chapter 4 touches on tort, contract criminal, and personal property law in relation to AI and robotics. Chapter 5 provides a high-level overview of concerns about automated decision-making.
Chapter 6 is dedicated to privacy, and discusses data collection and sharing, surveillance. It has brief sections on the use of data in scenarios like election campaigns, and defence, national security, and law enforcement. The last part of the chapter touches on data ownership, and the book takes the view that treating data as a ‘thing’ to which property rights attach does not fit easily with Australian law, because data is not an object of property rights under Australian law.
Chapters 7 and 8 discuss liability and insurance, and competition and consumer law respectively. Chapter 9 tackles the future of work. Perhaps Chapter 3 should have come after Chapter 10, which goes into modes of regulation, and examples of regulation targeting e.g., automated processing, bots, AI, and autonomous weapons systems.
Overall, I see ‘Artificial Intelligence, Robots and the Law’ as an introduction to the areas of AI and robotics. The language used throughout the book is generally simple enough for students and non-lawyers to comfortably understand the discussions within. There are references to Australian legislation and caselaw where the issue is jurisdiction specific. Otherwise, much of the book has international application. If you are already a specialist or have experience in AI regulation, I would suggest borrowing a copy for a read before deciding whether to buy the book, as the book does not go in-depth into any of the topics.
Darren Grayson Chng
About the book
Artificial Intelligence, Robots and the Law by Michael Guihot and Lyria Bennett Moses