Table Of Content

Trustworthy AI Trustworthy AI A Business Guide for Navigating Trust and Ethics in AI Beena Ammanath Copyright © 2022 by John Wiley & Sons, Inc. All rights reserved. Published by John Wiley & Sons, Inc. Hoboken, New Jersey. Published simultaneously in Canada. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per- copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8 400, fax (978) 750-4 470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6 011, fax (201) 748-6 008, or online at http://www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762- 2974, outside the United States at (317) 572- 3993 or fax (317) 572- 4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging- in- Publication Data is Available: ISBN: 9781119867920 (Hardback) ISBN: 9781119867968 (ePdf) ISBN: 9781119867951 (epub) Cover Design: Wiley Cover Image: Sunrise © titoOnz/Getty Images Contents Foreword vii Preface xi Acknowledgments xv Introduction 3 Chapter 1: A Primer on Modern AI 13 Chapter 2: Fair and Impartial 23 Chapter 3: Robust and Reliable 43 Chapter 4: Transparent 61 Chapter 5: Explainable 75 Chapter 6: Secure 93 Chapter 7: Safe 111 v vi Contents Chapter 8: Privacy 127 Chapter 9: Accountable 147 Chapter 10: Responsible 161 Chapter 11: Trustworthy AI in Practice 175 Chapter 12: Looking Forward 187 Notes 191 Index 199 Foreword T here are two kinds of organizations. Those that are fueled by artificial intelligence and those that will become fueled by AI. Eventually, all organizations public and pri- vate will be AI organizations. It is a twenty- first- century fact that efficiency, agility, competitiveness, and growth hinge on the suc- cessful use of AI. This is to the good, and the value is measured in trillions of dollars. The potential, however, comes with a caveat. We cannot cre- ate world- changing solutions and use AI for all the beneficial purposes we can imagine unless this technology rests on a firm ethical foundation. What we do today to align AI function and use with human values sets the trajectory for this transformative technology for decades to come. I have had wonderful opportunities to speak with government and business leaders from countries around the world. Ques- tions about AI ethics and trust are being debated in boardrooms, vii viii ForEW ord business halls, legislative chambers, and the public square. What I’ve learned is that, on the one hand, there is a growing awareness of how ethics and trust affect the use of AI. on the other hand, this awareness is causing some wariness about whether AI should be used at all. This speaks to valid concern about the technical function and social impact of AI, but it also reveals a more fundamental need on the part of organizations deploying AI. People want to trust AI, if only they could. The good news is we can get there, and it will take important decisions and actions to set not just individual businesses but the entire world on the path to trustworthy AI. More and more, leaders are realizing that we, as a global population, must act purposefully on AI ethics – and we must do so now. If we do, we can minimize the risks, maximize trust in the technology, and build toward the brightest future AI can facilitate. Some might speculate that prioritizing ethics and trust could impede innovation at the moment when the power of AI is finally brought to fruition. I submit that the reverse is true. The most powerful and useful AI innovations are those that align with our ethics and values. What we do not trust, we will not use. We need to be able to trust cognitive tools to move forward with AI. While the imperative is clear, the tactics and knowledge to address trust are somewhat less so. What ethics are relevant for AI use? What does it mean to trust a machine intelligence? Where are the hurdles, the pitfalls, and the great opportunities to build ethics into novel AI? The chorus of voices asking these questions and others like them is growing louder, and the refrain is the same: How do we get to a place where our AI tools are simultaneously powerful, valuable, and trustworthy? For better or worse, there is not just one answer to that ques- tion. There are many possible answers. Every organization oper- ates within a society, and communities, nations, and regions can