Law And Autonomous Machines: The Co-Evolution Of Legal Responsibility And Technology

261 Pages·2019·2.601 MB·English

by  Mark Chinen

Law and Autonomous Machines
ELGAR LAW, TECHNOLOGY AND SOCIETY
Series Editor: Peter K. Yu, Drake University Law School, USA
The information revolution and the advent of digital technologies have ushered in
new social practices, business models, legal solutions, regulatory policies, govern-
ance structures, consumer preferences and global concerns. This unique book series
provides an interdisciplinary forum for studying the complex interactions that arise
from this environment. It examines the broader and deeper theoretical questions
concerning information law and policy, explores its latest developments and social
implications, and provides new ways of thinking about changing technology.
Titles in the series include:
Transnational Culture in the Internet Age
Sean A. Pager and Adam Candeub
Environmental Technologies, Intellectual Property and Climate Change Accessing,
Obtaining and Protecting
Abbe E.L. Brown
Privacy and Legal Issues in Cloud Computing
Edited by Anne S.Y. Cheung and Rolf H. Weber
Intellectual Property and Access to Im/material Goods
Edited by Jessica C. Lai and Antoinette Maget Dominicé
The Legal Challenges of Social Media
Edited by David Mangan and Lorna E. Gillies
Digital Democracy in a Globalized World
Edited by Corlen Prins, Colette Cuijpers, Peter L. Lindseth and Mônica Rosina
Privacy in Public Space
Conceptual and Regulatory Challenges
Edited by Tjerk Timan, Bryce Clayton Newell and Bert-Jaap Koops
Autonomous Vehicles and the Law
Technology, Algorithms and Ethics
Hannah YeeFen Lim
The State of Creativity
The Future of 3D Printing, 4D Printing and Augmented Reality
James Griffin
Law and Autonomous Machines
The Co-evolution of Legal Responsibility and Technology
Mark Chinen
Law and Autonomous
Machines
The Co-evolution of Legal Responsibility and
Technology
Mark Chinen
Seattle University School of Law, USA
Cheltenham, UK • Northampton, MA, USA
© Mark Chinen 2019
All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system or transmitted in any form or by any means, electronic, mechanical or
photocopying, recording, or otherwise without the prior permission of the publisher.
Published by
Edward Elgar Publishing Limited
The Lypiatts
15 Lansdown Road
Cheltenham
Glos GL50 2JA
UK
Edward Elgar Publishing, Inc.
William Pratt House
9 Dewey Court
Northampton
Massachusetts 01060
USA
A catalogue record for this book
is available from the British Library
Library of Congress Control Number: 2019930684
This book is available electronically in the
Law subject collection
DOI 10.4337/9781786436597
ISBN 978 1 78643 658 0 (cased)
ISBN 978 1 78643 659 7 (eBook)
2
0
Contents
Preface and acknowledgements vi
Introduction viii
PART I THE RISE OF AUTONOMOUS
TECHNOLOGIES AND CURRENT LAW
1 The emerging challenge 2
2 Existing law and other forms of governance 23
PART II INDIVIDUAL AND GROUP RESPONSIBILITY
3 Individual responsibility 52
4 The legal and moral responsibility of groups 77
PART III REIMAGINING RESPONSIBILITY AND THE
RESPONSIBLE AGENT
5 Reframing responsibility 103
6 Altering the responsible agent 127
PART IV ETHICAL AI
7 Law-abiding machines and systems 147
8 Moral machines and systems 170
9 Machines and systems as legal and moral subjects 194
PART V CONCLUSIONS
10 Trigger events 224
Index 237
v
Preface and acknowledgements
The idea for this book emerged from the intersection of three areas of interest.
The first is of course the rise of artificial intelligence, its powering of machines
and systems, and the effects it might have on our lives. The second is the
legal responsibility of groups, including states, military organizations, and
multinational corporations. The third has to do with complexity theory and the
relationship between autonomous agents, human and artificial, and complex
systems. As artificial intelligence, autonomous machines and systems, and
intelligence enhancement have become more widely used, several of the issues
that emerge with the legal responsibility of groups are arising in debates about
how society in general and the legal system in particular should respond to
these emerging, increasingly sophisticated technologies, particularly if they
cause harm, and whether and how the law itself might change in relation to
those machines and systems.
It was useful to frame this exploration in terms of a possible trajectory for
the coevolution of autonomous technologies and legal responsibility, that is,
to examine how emerging technologies are interacting with existing systems
of responsibility, to identify particular trends, and to consider where those
trends might lead. As such, this study is not intended to be a complete review
of the existing legal doctrine as it might apply to autonomous technologies;
excellent works of that kind already exist. Of course, this book must describe
recent developments in the technology and the legal landscape to some extent,
but its purpose is to be more thematic and speculative in nature. At the same
time, over the course of writing the book, the project became more than tracing
a possible trajectory: it also became a chance for reassessment. The emergence
of artificial intelligence requires us to reconsider how we already address
harms and to ask whether the justifications for our existing systems of respon-
sibility are likely to remain persuasive as intelligent technologies become
a greater, perhaps ubiquitous part of our daily experience.
Thanks are due to many people. This book stems from an article published
originally in the Virginia Journal of Law and Technology.1 The article
benefited from the editorial work of the journal. Joshua Fensterbush, Kasie
1 Mark A. Chinen, The Co-Evolution of Autonomous Machines and Legal
Responsibility, 20 Va. J. L. & Tech. 338 (2016).
vi
Preface and acknowledgements vii
Kashimoto, and Maria Luisa Juarez y Hernandez provided valuable research
assistance. My faculty colleagues at the Seattle University School of Law gave
helpful comments at a presentation of an early version of the article at a faculty
colloquium. Much of the material that makes up Chapter 4: Complexity
Theory and its Meaning for Group and Individual Responsibility first appeared
as a book chapter in International Economic Law and African Development
(Laurence Boulle, Emmanuel Laryea, and Franzika Sucker, eds., Siber Ink,
2014). My thanks to the publisher for permission for use of the materials.
The book also has been improved by comments and conversations with Ben
Robbins and Stephen Wu, and by the suggestions given by the anonymous
peer reviewers who reviewed the initial proposal for this project. My adminis-
trative assistant, Laurie Wells, helped with the preparation of the manuscript.
All errors and omissions, however, remain my own.
My in-laws, Carl and Katherine Takushi, kindly allowed me the use of their
home where I wrote the bulk of the manuscript. Finally, I wish to thank my
wife, Ruby Takushi, and daughters, Maya and Grace Chinen, for their support
as this book was being written.
A note on citations: citations follow the Bluebook Uniform System of
Citation, with modifications to conform to the house style.
Introduction
The inevitable occurred on March 18, 2018, when an autonomous vehicle
owned and operated by Uber Technologies, Inc. struck and killed Elaine
Herzberg as she crossed a street with her bicycle. Her friends describe her as
just emerging from homelessness. A human driver was in the sports utility
vehicle, but the SUV was in autonomous mode when it struck Herzberg.1
Video footage indicates it was night and Herzberg was not in a crosswalk.
The footage also suggests the car did not apply brakes or otherwise attempt to
avoid hitting Herzberg. A preliminary report of the National Transportation
Safety Board revealed that the vehicle’s self-driving system first detected
Herzberg about six seconds before impact. As the car and Herzberg converged,
the autonomous system classified Herzberg as an unknown object, next as
a vehicle and then as a bicycle. Just 1.3 seconds before impact, the system
concluded that an emergency braking maneuver was required; however, such
maneuvers were not enabled while the vehicle was under computer control.
This was to prevent “erratic vehicle behavior.” The system was not designed
to alert the human operator.2
Others had died in accidents involving self-driving cars before this incident,3
but this was the first time someone with no voluntary connection to an auton-
omous vehicle had been killed by one. Immediately, discussions arose in the
comments sections of news reports and via Twitter posts about who should be
held responsible for Herzberg’s death. Was it the car? Uber? Herzberg herself?
The human driver who was there to override the SUV’s systems in an emer-
gency? (An accident report by the Tempe Police Department indicated that
1 Daisuke Wakabayashi, Self-Driving Uber Car Kills Pedestrian in Arizona,
Where Robots Roam, N.Y. Times, Mar. 19, 2018, at A1.
2 U.S. Nat’l Transp. Safety Bd., Preliminary Report: Highway, HWY18MH010
(2018), at 2 [hereinafter NTSB Preliminary Report].
3 A recent article from the popular press is Adrienne LaFrance, Can Google’s
Driverless Car Project Survive a Fatal Accident? Atlantic, Mar. 1, 2016, http://w ww
.theatlantic .com/ technology/ archive/ 2016/ 03/ google -self -driving -car -crash/ 471678/ .
The first death in a self-driving vehicle occurred on May 7, 2016. See Bill Vlasic &
Neal E. Boudette, As U.S. Investigates Fatal Tesla Crash, Company Defends Autopilot
System, N.Y. Times, Jun. 16, 2016, http://w ww .nytimes .com/ 2016/ 07/ 13/ business/ tesla
-autopilot- fatal- crash -investigation. html ? _r = 0.
viii
Introduction ix
the driver had been streaming a video program while in the car.4) The state of
Arizona for failing to adequately regulate autonomous vehicles? (I will return
to these facts and questions in Chapter 2.)
As that event was making headlines, another, perhaps more far-reaching
development was coming to light. This was the news that information from
millions of Facebook users had been acquired, without permission, by
Cambridge Analytica, which had in turn used that information to construct
behavioral models as it advised clients who wanted to frame and direct public
opinion.5 It was possible that as many as 87 million users had been affected.6
There was debate about whether such efforts to sway the public succeeded,
but at least some claimed that they had influenced important public decisions,
including Brexit and the 2016 US election. Again, questions were raised about
who should be blamed for the invasion of privacy, and again, there were likely
culprits: Cambridge Analytica and its principals, of course, but also Facebook
for collecting private data in the first place and then failing to protect it.
If nothing else, the loss of Herzberg’s life and the Cambridge Analytica
scandal announced that autonomous machines and systems and the harms they
can cause are no longer a distant prospect. The issue of how law can be used to
prevent them from harming others and how to assign responsibility if they do
is fast becoming ripe. Policymakers, industry leaders, and legal scholars have
seen this day coming for some time and there is a growing literature on the role
law can and should play as artificial intelligence and the technologies it powers
become a greater part of everyday experience.
This book contributes to that discussion by plotting a possible trajectory
for the relationship between the law and autonomous machines and systems.
That trajectory begins where it often does when something new appears on the
scene: communities use already existing legal doctrines and principles to com-
prehend and respond to it. This in turn leads to debate about how well suited
these doctrines and principles are to new developments and whether there will
be a need for significant changes in them.
4 Associated Press, Police: Backup Driver in Fatal Uber Crash was Distracted,
N.Y. Times, June 22, 2018, https://w ww .nytimes .com/ aponline/2 018/ 06/ 22/ us/ ap -us
-uber -autonomous -vehicle. html.
5 Matthew Rosenberg, Nicholas Confessore & Carole Cadwallader, How
Trump Consultants Exploited the Facebook Data of Millions, N.Y. Times, Mar. 17,
2018, https://w ww .nytimes .com/ 2018/ 03/ 17/ us/ politics/c ambridge- analytica- trump
-campaign. html.
6 Cecilia Kang & Sheera Frenkel, Facebook Now Says Improper Data Use
Affected 87 Million, N.Y. Times, Apr. 4, 2018, https://w ww .nytimes .com/ 2018/
04/ 04/ technology/ mark -zuckerberg -testify -congress .html ?hp & action = click & pgtype
= Homepage & clickSource = story -heading & module = first -column -region & region = top
-news & WT .nav = top -news.