The Techno-Legal Dynamics of Cyber Crimes in Industry 5.0 E-Book
173,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Fachliteratur
- Serie: Industry 5.0 Transformation Applications
- Sprache: Englisch
This book explores the core principles, technological advancements, and legal challenges of Industry 5.0’s digital transformation.
Industry 5.0 has enhanced the operational efficiency of the entire manufacturing process by incorporating multiple emerging technologies; however, high-tech cybercrimes have prompted legal scholars worldwide to rethink the fundamental principles of technology and law.
The Techno-Legal Dynamics of Cyber Crimes in Industry 5.0 shows how advanced technologies, such as artificial intelligence, the Internet of Things, and robotics, are integrated within manufacturing environments. It explores the intricate relationship between legal systems and technological advancements and addresses the rise of cybercrime following Industry 5.0’s digital transformation. Focusing on the interaction between technology and law, the book investigates current cyberlaw issues and solutions. It draws insights from diverse experts, including scholars, legal professionals, and industry leaders, emphasizing effective regulations to minimize cyber threat risks for Industry 5.0.
By adopting an international viewpoint, this book sheds light on various dimensions of nascent cybercrimes and legislative efforts worldwide aimed at governing them effectively.
Audience
This book should be read by legal scholars, lawyers, judges, legal and information technology researchers, cybersecurity experts, computer and software engineers, and students of law and technology. Regulators, policymakers, international trade specialists, and business executives should read it as well.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 724
Veröffentlichungsjahr: 2025
Ähnliche
Table of Contents
Cover
Table of Contents
Series Page
Title Page
Copyright Page
Preface
1 AI & IP: Ownership Rights in Industry 5.0
1.1 Introduction
1.2 Evolution of Artificial Intelligence System
1.3 Rights Upon the Fruits of AI: #TheConundrum
1.4 The Antidote to Conundrum: Work Made for Hire (WMFH) Model
Conclusion
References
2 Cybersecurity and Crime in Industry 4.0: An Analysis of Legal Aspects of Cybercrime
2.1 Introduction
2.2 Industry 4.0
2.3 Cybercrime in the Digital Economy
2.4 Cybercrime Statistics—Global Context
2.5 Cybersecurity
2.6 Legal Perspectives of Cybercrime and Their Penalty-Cases
2.7 Challenges of Cybersecurity
2.8 Conclusion
References
Weblinks
Case Links
3 Toward an Intelligent Cybersecurity System: The Role of Machine Learning
3.1 Introduction
3.2 Machine Learning Paradigms
3.3 Machine Learning Applications
3.4 Machine Learning Applications in Cybersecurity
3.5 Machine Learning in Cybersecurity—Few Cases
3.6 Issues and Limitations of Machine Learning in Cybersecurity
3.7 Future Directions
3.8 Conclusion
References
4 ATP the New-Age Threat Vector and Cyberattack Trends
4.1 Introducing Advanced Persistent Threats
4.2 New-Age APT Attacks
4.3 Cyberattacks Trends
4.4 Reconnaissance
4.5 Initial Compromise
4.6 Establishing a Foothold
4.7 Exfiltration
4.8 Consolidation
4.9 Covering Tracks
4.10 Conclusion
References
5 Online Privacy in Artificial Intelligence Algorithms: Ethical and Legal Impacts of Technological Development and Exposure
5.1 Introduction
5.2 Understanding Artificial Intelligence
5.3 Impact of Artificial Intelligence on Personal Data
5.4 General Data Protection Regulation and Artificial Intelligence
5.5 Right to be Forgotten: Legal and Ethical Considerations of Data Processing by an AI
5.6 Conclusion
References
6 Exacerbation and Combat of Cyberattacks: The Dual Paradox of Machine Learning
6.1 Introduction
6.2 Concept of Machine Learning
6.3 Technologies Involved in Machine Learning and Their Impact on Cybersecurity
6.4 Growing Role of Machine Learning in Cybersecurity
6.5 Applications of Machine Learning in Cybersecurity
6.6 Demerits of Using Machine Learning in Cybersecurity
6.7 Legal Framework for Cybersecurity in India
6.8 Conclusion
References
7 Hacking the System: A Deep Dive into the World of E-Banking Crime
7.1 Introduction
7.2 Literature Review
7.3 Combating E-Banking Crimes
7.4 E-Banking Crimes: Emerging Economies
7.5 Conclusion
References
8 Is Love a Crime-Decoding Cybercrimes in Online Dating and Risk Mitigation
8.1 Introduction to Online Dating
8.2 Cybercrimes via Online Dating
8.3 Exploring the Law of India
8.4 Laws Governing Dating Apps in India
8.5 A Ray of Hope—is Mitigation Possible
References
9 Critical Analysis of the Role of Intermediaries with Respect to Cybercrimes in Cyberspace
Introduction
References
10 Cybercrime and AI: Issues and Solutions
10.1 Introduction
10.2 Literature Review
10.3 Nature of Artificial Intelligence
10.4 Artificial Intelligence Tools for Cybersecurity
10.5 Legal Challenges
10.6 Comparison of Indian and International Law
10.7 Conclusion
10.8 Suggestions
References
11 The Illusion of Bitcoin: The Rise and Fall of a Revolutionary Cryptocurrency
11.1 Mystery of Bitcoin: An Overview
11.2 The Rise and Fall of Bitcoin
11.3 The Illusion of Bitcoin
11.4 Bitcoin and Cybersecurity
11.5 Measures to Enhance Bitcoin Security
11.6 Case Studies of Bitcoin Security Breaches
11.7 Bitcoin and the Financial Industry
11.8 Social and Political Implications of Bitcoin
11.9 Bitcoin’s Role in Crime
11.10 The Future of Bitcoin
11.11 Conclusion
11.12 Way Forward
References
12 Cybercrime Against Women in India: Challenges and Possible Solutions
12.1 Introduction
12.2 History and Origin of Cybercrimes
12.3 Literature Review
12.4 Reasons for the Growth of Cybercrimes Against Women
12.5 Different Types of Cybercrime Against Women
12.6 Implication of Cybercrimes on Women
12.7 Legal Provisions Against Cybercrimes
12.8 Provisions for Cybercrimes Under IT ACT 2000 [20]
12.9 Landmark Cases in Cybercrime Against Women
12.10 Shortcomings in the Indian Legal System
12.11 Suggestions
Conclusion: A Way Forward
References
13 Emerging Legal Challenges in the Artificial World of Metaverse
13.1 Introduction
13.2 The Ideas Behind the Metaverse
13.3 The Application of the Metaverse in Different Industries
13.4 Issues and Challenges after Introducing the Metaverse
13.5 Approaching Metaverse Problems from a Global Perspective
13.6 Conclusion
References
14 Cyberterrorism: In an Era of Information Warfare
14.1 Introduction
14.2 The Surge in Cyberterrorism
14.3 Background
14.4 Methodology Used
14.5 Effects of Cyberterrorism on Society and Government
14.6 Techniques and Methods Used for Cyberterrorism
14.7 Countries Affected the Most by Cyberterrorism
14.8 Ukraine and Russia War and Cyberterrorism
14.9 Combating Cyberterrorism
14.10 Conclusion
References
15 Freedom of Speech and Expression on Social Media—A Comparison of India and China
15.1 Introduction
15.2 Review of Literature
15.3 Online Speech Regulations—India vs. China
15.4 Intermediary Responsibilities—India vs. China
15.5 Conclusion
References
16 Environment and Cybercrime in the World: Mitigating the Impact of Cybercrime on the Environment
16.1 Introduction
16.2 Impact of Cybercrime on the Environment
16.3 Mitigating the Impact of Cybercrime
16.4 Future of the Environment and Cybercrime
16.5 Conclusion
References
17 Face Recognition Based on Smart Attendance System Using Python
17.1 Introduction
17.2 Methodology
17.3 Face Database
17.4 Implementation of CNN and SVM Algorithm
17.5 Features and Benefits Features of the System
17.6 Conclusion
References
Index
End User License Agreement
List of Tables
Chapter 7
Table 7.1 Comparison of e-banking cybersecurity challenges and trends in India...
List of Illustrations
Chapter 2
Figure 2.1 Cybercrimes trend
Figure 2.2 Size of the cybersecurity services market in India from 2018 to 202...
Chapter 5
Figure 5.1 Deletion in MySQL 20 [28].
Chapter 13
Figure 13.1 Global metaverse revenue by CAGR.
Chapter 16
Figure 16.1 Share of primary energy from fossil fuels in 2019
Chapter 17
Figure 17.1 The operating process of the attendance system.
Figure 17.2 Flow diagram.
Figure 17.3 Data enter.
Figure 17.4 Face-to-face analysis.
Figure 17.5 Capture dataset.
Figure 17.6 The face detection using Haar-cascade.
Figure 17.7 The face features local representation.
Figure 17.8 LBPH recognition.
Figure 17.9 Face recognition using CNN.
Figure 17.10 The framework of the whole training system.
Figure 17.11 The framework of testing.
Figure 17.12 Structure of CNN.
Figure 17.13 Pre-training and not use pre-training.
Figure 17.14 The proposed system.
Guide
Cover Page
Table of Contents
Series Page
Title Page
Copyright Page
Preface
Begin Reading
Index
WILEY END USER LICENSE AGREEMENT
Pages
ii
iii
iv
xv
xvi
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
393
394
395
396
397
Scrivener Publishing100 Cummings Center, Suite 541JBeverly, MA 01915-6106
Industry 5.0 Transformation Applications
Series Editors: Dr. S. Balamurugan (sbnbala@gmail) and Dr. Sheng-Lung Peng
The increase in technological advancements in the areas of artificial intelligence (AI), machine learning (ML) and data analytics has led to the next industrial revolution “Industry 5.0”. The transformation to Industry 5.0 collaborates human intelligence with machines to customize efficient solutions. This book series covers various subjects under promising application areas of Industry 5.0 such as smart manufacturing, intelligent traffic, cloud manufacturing, real-time productivity optimization, augmented reality and virtual reality, etc., as well as titles supporting technologies for promoting potential applications of Industry 5.0, such as collaborative robots (Cobots), edge computing, Internet of Everything, big data analytics, digital twins, 6G and beyond, blockchain, quantum computing and hyper-intelligent networks.
Publishers at ScrivenerMartin Scrivener ([email protected])Phillip Carmical ([email protected])
The Techno-Legal Dynamics of Cyber Crimes in Industry 5.0
Edited by
Gagandeep Kaur
Associate Professor in Law, School of Law, UPES, Dehradun, Uttarakhand, India
Tanupriya Choudhury
Professor, School of Computer Sciences, University of Petroleum and Energy Studies (UPES), Dehradun, Uttarakhand, India
and
S. Balamurugan
Director of Intelligent Research Consultancy Services, Coimbatore, India
This edition first published 2025 by John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA and Scrivener Publishing LLC, 100 Cummings Center, Suite 541J, Beverly, MA 01915, USA© 2025 Scrivener Publishing LLCFor more information about Scrivener publications please visit www.scrivenerpublishing.com.
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, photocopying, recording, or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.
Wiley Global Headquarters111 River Street, Hoboken, NJ 07030, USA
For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.
Limit of Liability/Disclaimer of WarrantyWhile the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchant-ability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials, or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. 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. 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.
Library of Congress Cataloging-in-Publication Data
ISBN 978-1-394-24214-6
Cover image: Adobe FireflyCover design by Russell Richardson
Preface
Industry 5.0 signifies the most recent advancement in the progression of industrial technologies, marked by a seamless fusion of cyber-physical systems, artificial intelligence (AI), and the Internet of Things (IoT). This new era has significantly enhanced operational efficiency across manufacturing processes through the adoption of a variety of cutting-edge technologies such as blockchain, cloud computing, robotics, and AI. The hallmark of Industry 5.0 is the merging of tangible and digital realms to forge interconnected frameworks that bolster automation capabilities, improve communication channels, and refine data analysis procedures. These cyber-physical systems afford manufacturers the ability to monitor and manage physical operations in real-time while advanced AI algorithms facilitate more effective decision-making processes. The ultimate objective of Industry 5.0 is to establish intelligent factories equipped with smart production systems that render manufacturing environments highly responsive and cognizant. By doing so, it ensures that this iteration maintains all critical concepts from the original text while offering alternative expressions for those ideas without compromising clarity or contextual relevance.
As Industry 5.0 ushers in a new era of exceptional efficiency and breakthroughs, it simultaneously presents fresh challenges, especially concerning the security of our digital infrastructure. Among these emerging issues is the rise of cybercrime—a term that captures an array of illicit or harmful actions involving computer systems, online networks, and information technology. In the aftermath of Covid-19, criminal activities in the digital domain have surged significantly. Law enforcement agencies may adapt to combat these crimes; however, resourceful and shrewd criminals consistently devise novel methods to perpetrate offenses using sophisticated and advanced technological means. Cybercrime today cannot be pigeonholed into a single category; rather it represents a spectrum of illegal acts that are compelling legal scholars to revisit core tenets within jurisprudence. The high-tech landscape fostered by Industry 5.0 exposes itself to countless cyber threats that jeopardize its integrity.
In the current era of advanced industry, the crucial role played by the interaction of technology and legal structures in comprehending and tackling cybercrime cannot be overstated. With operations increasingly driven by Industry 5.0’s interconnectivity and digital transformation occurring at a rapid clip, the potential impact of cyberattacks is unprecedentedly broad, posing significant challenges for manufacturers and their supply chains that may be ill-equipped to handle such threats. These risks do not recognize national borders; they are international in scope, impacting entities across countries.
The text within this volume delves into the intricate relationship between legal systems and technological advancements on both domestic and global stages. The contributors have meticulously investigated contemporary cyberlaw issues to address areas lacking adequate legal provisions while incorporating insights from scholars, lawyers, educators, budding jurists, and industry experts. By adopting an international viewpoint, this book sheds light on various dimensions of nascent cybercrimes as well as legislative efforts worldwide aimed at governing them effectively.
The EditorsDecember 2024
1AI & IP: Ownership Rights in Industry 5.0
Pulkit Mogra
Faculty of Law, Univesity of Ottawa, Ottawa, Canada
Abstract
Creativity and expression are no longer a forte of a human being. Advanced algorithms, also known as AI, are capable of creating content of their own such as painting and composing music. In the era of Industry 5.0, Google’s AI company, DeepMind, has created software that can produce novel music sounds and unique images. If such things had been created by humans, they would have been subject to protection under copyright laws but since machine lacks the characteristics of humans, the question arises about who owns the copyrights for the content.
US circuit court has recently held in the case of Naruto et al. vs David Slater that animals, other than humans, cannot sue for copyright protection. Furthermore, WIPO member countries enacted a law that states non-humans are not subjected to protection under IP laws. Furthermore, it depends a lot upon the interpretation of courts for originality requirements under authorship and if it requires creative inputs from humans. In India, to get protection under copyright, one must prove creativity in addition to variation from previous works.
Another issue that this paper reflects on would be the determination of several possibilities for assigning authorship where an object has been designed by AI. This paper will analyse the claims of the stakeholders such as programmers, users, and AI software itself to determine who has the greatest claim towards ownership of products and further possibilities for assigning authorship where an object has been designed by AI, thus analysing the question of “who is the author” of the works created by AI.
This piece contributes to the literature on conundrums arising in the Industry 5.0 era, which on hand is striving hard to develop technologies in assistance of AI but is also facing issues on who owns the assets and liabilities arising out from the labour of AI. With the press release of the Japanese government that plans to create a legal framework, especially for the works created by AI to protect copyrights on novels, music, and other works, it becomes a relevant question if laws, including international treaties, should redefine the term “authorship” and include non-legal entities as well under the umbrella or should develop a new legal framework for AI.
Keywords: Artificial intelligence, intellectual property law, copyrights
1.1 Introduction
Artificial intelligence (AI) system–driven robots and other systems with AI technology possess the capabilities of creating independent original works, i.e., without the intervention of human beings. The modern AI system consists of ten characteristics, i.e., (i) autonomous, (ii) free choice, (iii) innovative, (iv) independent, (v) unpredictable, (vi) intelligent and rational, (vii) efficient, (viii) capable of evolving and learning, (ix) accurate, and (x) goal-oriented, all such features together make a system intelligent [1]. That raises many legal questions: Are these works copyrightable as original artwork? Who owns the copyrights, along with the bundle of ownership rights such as royalties, licences, right to get remedies in case of infringements, and so forth? Let us imagine that an AI-driven computer system generates an image of a spaceship based upon a few inputs by the user, such as “company logo” and typing terms such as “Space,” “Ship,” or “Pyramid.” Further to that, an image is being generated; now, does a user have a right to sell the image without the programmer’s consent? Some possible answers to this question can be that the right may be assigned to the programmer, user, both user and programmer, AI system itself, both user and AI system, both programmer and AI system, or nobody.
Although most of the principles have been in favour of awarding the rights to users, they have evolved since the 1800s when technology was merely a tool for the users such as typewriters and cameras [2]. However, with time, computer programmes became more and more intelligent and sophisticated and, hence, started to display their own set of creativity without much intervention from users. Therefore, weighing the programmes in the same set of tests is no longer relevant, and hence AI-based computer programmes are required to be tested from the new lenses with the broadened category of ownership rights. It seems challenging to transpose it in combination with the norms of existing legal systems connected to the Droit d’Auteur tradition, where the creation must result from an effort of human intellect.
This paper deals with finding out the best possible solution amongst the variety of existing suggested frameworks, which would allow the rights and liabilities to be distributed amongst programmers, owners, and users of AI, and at the same time also promote the innovational incentive for the society.
The first part of the paper briefly explains the definition of an AI system, classifying it under “Strong” and “Weak” AI systems. Furthermore, it explains how the competency of an AI is determined through a test suggested by Prof. Alan Turing in 1950 in his paper “Computing Machinery and Intelligence” [3]. This part also highlights the creativity of AI systems and what it means when we use the term “creative” in the AI context. Traditionally, creativity has been seen as a component of human intellect and copyright law in the EU, UK, and many other jurisdictions that state the importance of creativity explicitly in granting up copyrights.
The second part deals with the contemporary issue of “Who owns the rights in the works created by AI systems”? This part lays down and analyses the US and EU’s legal position on the subject matter along with the Macque and Fiest Publications case and the stance of the Berne Convention upon copyrightability. Furthermore, this part examines the stance of various stakeholders that may claim ownership of AI-developed products or services, such as the programmer, AI itself, the user, or the public domain, and also lays down a test where an AI can be qualified as an independent AI and thus concluding with the recent developments.
Lastly, the third part of the paper suggests a possible solution by examining the principles of employer-employee relationships and licencing. Also, this part mentions the advantages of the “Work Made for Hire” along with the “Licencing model” and how this model creates a win-win situation for every stakeholder. This brings an important discussion for stakeholders of Industry 5.0, a term coined by the European Commission Report on “Industry 5.0, Towards a sustainable, human-centric and resilient European industry”, which states the importance of technologies such as AI that contribute towards unprecedented opportunities [4]. Whilst AI or robotics allows for workplace innovation and capitalisation on human-machine interaction, it becomes a pertinent question to discuss who may own the rights and liabilities arising out of this interaction. This model contributes towards easing the issue and provides a plausible solution to the questions raised in the paper, thus concluding with remarks on the model and also highlighting the innovative trends from countries such as Japan that have sought a pioneering approach to deal with this contemporary issue.
1.2 Evolution of Artificial Intelligence System
The term “Artificial Intelligence”, a neologism term, was first coined during the Dartmouth Conference where John McCarthy in 1956 at MIT defined AI as the science of developing machines along with the intelligent computer programme [5] (Sejnowski, 2018). However, AI with time has evolved into multi-folds, and so has the definition. In today’s era, there exists no universal definition of AI; linguistically, if we analyse the terms, it leads us to the following conclusion:
Artificial, in a literary sense, means something unreal that has been created by humans through various scientific techniques
[6]
.
Intelligence infers the ability to live organisms to undertake a decision-making process when presented with sets of data or situations, through planning, reasoning, analysing, and understanding
[7]
.
Combining these two terms, the simple explanation of AI is a machine that is capable of simulating human behaviour, as also introduced by Norvig and Russell, for a scale to measure the ability of non-human intelligence, in terms of its ability to replicate human skills and mental smartness such as recognising of patterns, learning through experiences, and reasoning [8].
In general terms, AI can be classified on the grounds of their type of approach towards technical implementation or their application as per the situation. The prominent type of AI can be classified under these categories, keeping in consideration their intended and displayed intelligence:
General AI: Also known as artificial general intelligence, an AI that can be referred to in a manner that it can function, think, and act similar to how a human brain might act and it would be difficult to distinguish between the acts of both
[9]
.
Narrow AI: Unlike general AI, narrow AI is not associated with duplicating human intelligence activities but works towards being associated with the attempts to create programmes that act as a supplement to a given task. Currently, most of the AIs that are being in use are classified under weak AI, and its use in specific applications can be witnessed under language translations, playing games where AI acts as an opponent, image recognition, and so forth
[9]
.
The “general AI” is the one that we shall be debating upon in this paper, as its autonomy and thought process leave us with the questions of rights and liability of its results. The general AI is the one that meets the test that was proposed by Turing in 1950 to determine if an AI can be equated to a human. Turing developed a test where participants were asked to interact with a machine in a text-only format [3]. Participants were required to indicate if they were having a conversation with a machine or a human. The test proposed that if the responses received from the participants were indistinguishable in comparison to how a human may respond, the machine would be classified under general AI [10].
Before we understand the challenges that general AI possesses, firstly, it becomes essential for us to understand how autonomous AI systems can produce creative work in a similar manner; if in case they had been created by humans, they would have been copyrightable [1]. Therefore, the features of AI can be classified into ten categories, which AI uses to independently develop the work that is useful and autonomously created [11].
Creative: AI machines operate in a manner that is more than capable of just copying works from other sources; they can operate creatively and produce original works. In terms of IP protection, this is one of the essential features for granting copyrights
[11]
.
Autonomous: A system can be classified as autonomous if it can complete complicated tasks without any intervention through external sources
[11]
.
New and unpredictable results: AI systems work upon an algorithm that is capable of producing new and unpredictable results such as creating a new painting, rather than copying from the existing works. Just like how humans are exposed to colours, shapes, and sizes, similarly, the AI systems upon their exposure to the existing knowledge forms, if they can produce fresh and unpredictable results, make themselves distinguishable
[11]
.
Data collection and communicating with the data: Communication is a necessary feature, where an AI must be able to independently communicate with the data of other AI systems, such as in the case of making choices during the collision of two autonomous cars.
Capable of learning: An AI system must be one that continuously is able to evolve itself through data processing and feedback, such as in the way SIRI has been developed.
Rational: This is one such capability that an intelligent system must possess, i.e., how to use the data at the optimum level to decide and make decisions and maximise the likelihood of achieving the goal
[8]
.
Evolution: AI system tends to evolve continuously as they interact with new environments and characteristics every day, and the machine might constantly find new patterns. In this sense, this feature of evolvement becomes a core feature of AI
[12]
.
Efficient: AI systems efficiently reflect their accuracy in nature as they process a massive amount of data rapidly that is far beyond the abilities of humans
[13]
.
Free choice: AI systems can choose between different alternatives and can derive the best results from the possible alternatives, such as driverless cars choosing the best route as per different traffic conditions
[14]
.
Goal-oriented: Functions of AI systems are very target-specific, such as drawing, painting, driving, composing, or reading
[14]
.
AI systems that develop artworks or perform any other tasks that are entirely autonomous in nature possess all these ten features to a certain extent. Once we comprehend this idea that AI systems through the abovementioned ten features can create results autonomously and independently in nature, one can understand that the rights available under the copyright laws cannot be considered to be exclusively available for human authors only, and therefore, traditional copyright laws and theories would not be applicable either on the given scenario. It becomes imperative to find the solution through various alternatives of how these rights can be distributed in a way that neither the innovation of the society nor the judiciary suffers from the excess infringement cases due to the ambiguity of laws.
Under copyright laws, the eligibility of subject matter flows from the concept that the work must be more than imitation, and it also requires that the creation must be independent of another original work [15]. Similarly, on the patent side, the definite requirement is “inventiveness” and must not have a resemblance to the prior art [16]. Since both copyrights and patents have a baseline of being independent of the previous work, therefore the test for AI subject matter comes down to the same test as in common law being applied to the natural person, where an AI must be able to reflect its capabilities of an independent creator. In simple words, an AI that would be providing results through a step-by-step process of algorithm would not qualify under the test as compared to the AI that shall be developing products through its own capabilities and learning as we discussed above in the ten features and the requirement of passing the Turing test.
Some of the examples would be:
Push Button Berta- would not be eligible, where the music was developed as an output to a simple algorithm, where the work was developed by the programmer, and the AI was merely providing variation
[17]
.
Google’s magenta project- would be eligible as Google’s AI system relies upon deep learning and neural networks for creating the original music
[18]
.
IBM Watson, Chef- would be eligible as Watson’s AI system relies upon inductive reasoning to develop food recipes that are non-obvious
[18]
.
Therefore, with the given examples and based upon the Turing test, it becomes clear that AI must be the reason for creativity as compared to just a mechanical device that is under the authority of the inventor. In 1884, in the case of Burrow-Giles Lithographic Co. v. Sarony, the U.S. Supreme Court clarified through reasoning over the dispute of holding the copyrights in the photograph that merely the mechanical process of capturing a picture does not grant copyrights to a camera, the creativity is the one that is an essential feature because of which the picture has resulted in a reality [2]. The creativity is the one that lies with the photographer and not with the camera, and the ownership lies with the originator for determining the authorship.
Similarly, it becomes quite evident that merely using the technology during the process of inventing does not make technology an inventor such as using Microsoft Word to keep the notes of the invention. The debate arises when the de minimus role of the human is involved in the process. If the AI system had adjusted the lights, selected costumes, and then captured a photograph, the debate would arise upon the question of, in this case, whether “mental conception” was enough to grant the copyright or not. Therefore, the test would come down to the role of AI if it has been creative or mechanical.
Hence, when an AI that can satisfy the tests and seems to be responsible for creating a work of art, makes a necessary decision for completing a service or developing an invention under its authority, the debate shall arise over who would own the IP rights for the fruits of the AI. The eligible parties may be the AI itself, the programmer, the user, or the public domain. The next part of the paper deals with the issue that if the AI passes the tests and proves that the products, artworks, and services are resultant because of its creativity and can imitate human behaviour, who shall be responsible for the IP rights and what would be the best possible solution to distribute the rights in a way that does not result into conflicts between several parties.
1.3 Rights Upon the Fruits of AI: #TheConundrum
In order to solve the conundrum of who may bear and enjoy the fruits of AI, i.e., IP rights over the creation of a product developed by AI, without human intervention, the concepts of “Author” and “Work of Authorship” hold the utmost importance. In most of the legal systems where there is a requirement for creativity, one may argue that only humans may be qualified to hold authorship, especially in civil law countries creativity is closely connected to a human and is generally seen as an emanation of the personality of the author.
Legislatures in several jurisdictions such as the UK, New Zealand, Australia, India, and South Africa have explicitly stated in their copyright acts that the rights over computer-generated works shall belong to the person who undertook the necessary arrangements for the creation of the work (UK Copyright, Design and Patents Act of 1988, C. 48, C.1., S. 9(3); South Africa Copyright Act of 1978, No. 98, (1978); Copyright Act of New Zealand of 1994, 1994/143, S. 5(2)(A).
In the United States, the Supreme Court in the case of Community for Creative Non-Violence v Reid defined an author as a party who creates work and the person who translates an idea into a tangible, fixed-expression community [19]. In Aalmuhuammed v. Lee, the court further made it clear that the author is likely to be a human [20]. Furthermore, after keeping a consistent view of an author being human, the Supreme Courts provided further additions to the modern copyright law through the case of Fiest Publications, Inc. v. Rural Telephone Service Co. Inc., where the judges stated that a copyrightable work must possess the spark of creativity along with the requirements of originality [21]. Therefore, originality and creativity have been critical aspects of authorship, and if we add up the requirements that the copyright can be held only by humans, in that case, AI can never be considered to be eligible for being an author.
Although not directly related to AI, the recent monkey selfie case, Naruto vs Slater, gave a comprehensive understanding of how courts examined the debate about non-human’s right to apply for copyright protection [22]. The case was dismissed by the court’s reasoning that the copyright act does not grant animals and non-humans, the right to hold copyrights. The court examined that under the United States law, there is a demand for “human creativity” and only that work is copyrightable, which has been created by humans [22].
In a similar sense, the paradoxical question that arises here as per the current scenario of law is that, if two works are identical, one has been created by a human and the other has been created by an AI system, will the treatment under copyright law be different or same? The answer is that the treatment would be different, where the autonomously created work by the AI system would face rejection under the copyright law. The status of law currently requires that there must exist legal personhood of the copyright holder, which an AI lacks [6]. However, there exist some ways by which this problem can be solved; since the law requires legal personhood, non-natural persons such as companies may be able to hold the rights through the work-for-hire doctrine, just in a way where an employee acts under the scope of employment and the owner is the one that holds the rights in his work. Some of the examples include works such as the drafting of a newspaper article by a journalist or software developed by the programme under the scope of his employment.
B. However, who is the author?
The answer to the above question is a bit complex when it comes to defining an author for AI-generated works as the answer revolves around factors such as striking the balance between economic interest and justification of providing the rights. Providing the rights to programmers and users, which are two of the parties that are involved in computational creation, also provides us with a less controversial, easy solution. Also, this goes along with the theory of providing copyrights only for creativity through humans. Nevertheless, in the case of AI, there seem to be some shortcomings under this approach due to the inputs of multiple parties up to a limited extent as well as the autonomy of AI itself.
There are three parties that may claim copyright over the works generated by AI, i.e., programmers, owners of AI such as companies and investors in AI, and thirdly, end users. Whilst determining the best possible solution for the authorship, it is required to be kept in check the total societal benefit whilst granting the copyrights to any of the parties. Under this section, we shall assess the claim of each of the parties in order to find out which of the parties can establish the closest claims.
1. Users of the AI System
The conventional way of how the courts have dealt with this is by granting ownership to inventors and artists who have created a product using machines [23]. Just as CONTU in its report stated in 1978, machines like typewriters, cameras, and computers are mere instruments that are only capable of functioning through the direct or indirect involvement of humans [24]. However, 1978’s analysis of the machines cannot be weighed on the same scale as today’s AI. The basis of courts providing human users as the owners of such outputs is that the courts have assumed that human creativity and originality are something that has led to the creation of work. However, AIs that can carry over the entire process of developing a product or an artwork let the old notions of originality and creativity qualify. As also discussed in the ruling of Sarony, an author is the one that is maker and originator and one who is responsible for completing the work. In the case of claims by “Users”, a user in the works generated by AI systems provides minimal or no guidance at all for the creation of the work, and hence IP ownership of the user would render it void [2].
Another shortcoming for the users as owners of the work is the issue of infeasibility if they are being granted the copyrights. Allocation of authorship or rights to users would eventually lessen the incentives amongst programmers to develop AI systems as the users would lead to “free riding” issues over the skills and labour of programmers.
2. The Programmer of the AI System
One of the most convincing arguments for granting the IP to the programmer is due to the fact that the programmer has invested his or her time, creativity, and energy in developing the AI system [25]. However, there also lies an argument that a programmer programmes an AI system through step-by-step processes and instruction through a developed algorithm. The algorithm that a programmer develops is already eligible for copyright protection. Furthermore, the programmer in the case of autonomous AI systems does not have control or predictions over what kind of work an AI system may create. The programmer merely develops the potentiality for an output to be created but not the actual output [26]. The creativity and originality of work are entirely a result of an AI system. Also, AI systems generally work on experiential learning just as humans do [27]. Another pertinent issue that lies with the programmer is economic issues. If a programmer of an AI system is the one who retains all the IP rights, in that case, the other parties such as users would be disincentivised, who may be purchasing the AI system to use it for monetary purposes, as they would not have the claim and the IP rights on the works of AI system [26]. For example, if IBM (programmer) is the owner of all IP rights for the pharmaceutical products created by the AI system Watson, the users may not buy Watson at all since that would compromise their ability to grant a patent in their name. Thus, due to the lack of justification for providing rights to the programmer and the inapplicability of creativity and originality tests over the efforts and role of the programmer for the output developed by AI systems, it would not be justified to provide the rights to programmers for the AI that has been algorithmically being programmed by the programmer.
3. AI System as the Owner
One of the critical reasons why IP rights, such as copyrights and patents, came into being was to provide economic incentives to the parties involved in the process of developing an AI system. Prof. Wu has argued that the assignment of authorship to AI systems might be permissible under the following circumstances.
If AI can decide upon when and where to produce works in the future; and
There is entirely no intervention by humans, and the algorithm acts independently; and
AI produces works that cannot be anticipated in nature
[26]
.
An AI machine can only be deemed as an author after it develops autonomy along with self-awareness similar to humans. At the current stage of technology, this seems hardly achievable; although some forms can mimic human creativity, complete autonomy is still yet to come. Also, as discussed above, machines do not require incentives to create any of their work, in such a scenario providing authorship and IP rights to AI systems themselves would leave stakeholders unrewarded. The programmers shall be deprived of all the rights that they have associated with their labour and creativity which in turn would lead to affecting the overall innovation of the society as this would discourage the programmers as well as users from developing and to use AI developed algorithms and AI machines, respectively [13].
4. Public Domain
Placing the rights of works created by the AI system into the public domain would mean that no one would own the rights to the works generated by the AI system. This solution theoretically may be consistent with the assumption that since the authors of the works are required to be humans, the creative spark is required to be sufficient enough. In the case of autonomous AI, since the role of humans is minimal, then so is the human creativity in the output created by AI. Therefore, the public domain could be one such area that would be undisputed for ownership rights.
However, this raises a bigger question that if no protection would be granted to AI, the efforts and investment that have been provided in the development of such a system would also be limited. IP rights would provide the incentive to the parties such as programmers, investors, and users to keep innovating and researching for future AI systems [28]. The above policy-related considerations suggest that eliminating the copyright protections and bringing the creation of AI systems into the public domain would do more harm than good since if there is no financial incentive available for launching a product in the market, then a product shall never find the relevant market.
In this research, I would like to argue that none of the actors mentioned above is entitled to own the AI-generated works. Due to the features of AI, as discussed in this paper such as autonomous, creative, unpredictable, evolving, and so forth, disqualifies all the above players from claiming any direct ownership of the AI-generated works. Furthermore, it must be noted that since there are too many players that are involved such as government, programmer, user, data providers, feedback providers, owner, and the public itself, none of the above players can be classified as the primary contributor for the work created by AI systems.
Therefore, to carve out the solution to this existing problem I would like to propose a model in the next part that shall aid us in striking a balance between the different entities along with not compromising on the overall innovation and growth of the society. This model would acknowledge the perception of automated, autonomous, and advanced AI systems along with providing accountability and control over the legal entity.
1.4 The Antidote to Conundrum: Work Made for Hire (WMFH) Model
The work made for hire is a model that considers an AI system as an employee that has been employed by a firm or a human, which legally operates the AI system. In a conventional sense, the person who owns the copyrights has exclusive rights to authorise others to prepare derivative work and distribute or reproduce the work. The problem with an AI system is that it would be unable to enforce the rights by itself, as an AI system would be unable to sue another AI system or an entity by itself nor it would be able to transfer those rights onto others to sue on them on its behalf. Under WMFH, the situation would be very similar to how a cameraman being a creative person shoots the whole film, but it is the director who holds the copyrights in the film. The situation of AI systems is quite similar to that creative cameraman. In the case of Goldstein vs California, the Supreme Court held that the requirement of authorship includes “physical rendering of fruits of creative intellectual or aesthetic labour”[29]. The court further reasoned that for a computer to generate any kind of artistic work, it may require input from the user [19].
Under the current set of laws, ambiguity arises over what must be done for the AI systems that can completely work on themselves, as the law still lacks the recognition of AIs as a natural person.
The Work Made for Hire Model is one such option to find the solution to this ambiguity.
A. The WMFH Legal Doctrine
The WMFH doctrine provides employers, contractors, or individuals with the right to have copyrights for the works that have been created by their employees, as the employer shall be treated as the one who has commissioned the work [19].
The U.S. Supreme Court in the case of Comm. for Creative Non-Violence v. Reed whilst addressing the work made for hire doctrine stated that the work made for hire shall only qualify if there exists an agreement between the parties that specifies that work was made for hire [19]. Furthermore, to establish the employer-employee relationship, the factors must fall under the three broad categories such as:
Control of employer over the employee, i.e., control of an employer over what kind of work has to be created.
Control of employer over the work, i.e., to determine how the work shall be done and means to do the work.
Conduct of employer, i.e., the related business of an employer and the benefits to an employee
[19]
.
Though the factors have not been stated as exhaustive in nature, the court left an ambiguity over what factors must be present to establish a relationship under WMFH. The rationale behind this doctrine is to create an incentive for the employer on whose direction the work has been prepared. The employer shall be the one who owns both rights and liability for the outcomes. In most cases, the users that shall be operating the AI system and providing the directions to the machine can be granted the copyrights. Also, for the users, the role is justified because the users also undertake the financial risk of buying and hiring the AI system and further supplying the required materials for the production of products, artworks, etc.
Furthermore, from this point of policy, it makes absolute sense to incentivise people to use the AI system as that shall also promote growth, scientific innovation, and the overall welfare of society [29]. Therefore, through WMFH doctrine, it shall allow a user to legally sell, use, or distribute the works that have been created by the AI system as well as it shall also place accountability for such products on a person under tort law and criminal law [30]. In the WMFH model, the AI system shall be recognised as a creative employee and will have its own legal personhood, just like a company is an artificial living entity. The autonomous AI system will be considered a creative author of a work, and when the AI system is acting autonomously, it will be shielded under the WMFH doctrine.
B. Licencing of Rights with Programmers
Since programmers act as a fundamental link in designing algorithms and creating and maintaining the AI systems, incentivising them becomes a pressing issue that must be addressed.
In the WMFH model, AI shall be named as an author for the works and will act as an employee to the user, and the user shall act as an employer with the rights over the works created by the AI system. The programmer’s recognition and incentives for his labour can be awarded through the assignment scheme as mentioned hereunder:
Explicit licencing agreement: under the licencing agreement, the AI system can be licenced by the user from the programmer, and the ownership of IP rights can further be transferred to the users. For example, IBM can enter into a licencing agreement with the user for its AI Watson for using the product and all IP rights can be transferred to the user.
Implicit agreement: Under the following, the AI can be licenced for being employed to invent, i.e., for a specific purpose for using the AI, and the associated IP rights can be transferred to the user for which the AI has been licenced.
In both situations, the programmers get recognition and monetary benefits through licencing the AI systems in order to serve as employees to the users. Since users shall be within the Work Made for Hire model, the users shall act as employers, and AIs shall serve them as employees. Furthermore, the rights and liabilities would automatically be transferred to users, hence solving the dilemma of who would be accountable for the rights and liabilities of AI systems.
Since the programmers often are unaware of the results of the algorithms because of the autonomy of AI systems, it would be unjustified to grant them rights for the works in which their creativity and predictability of the results generated are minimal. Hence, the WMFH model, along with the licencing scheme of AI systems from programmers, serves as a boon to the existing conundrum over ownership of rights and liabilities for the works generated by AI systems.
C. Advantage of WMFH Doctrine Along with the Licencing for Programmers
The WMFH doctrine brings along a number of benefits such as:
The model recognises an AI system’s creations as both independent and creative and imposes the same rules and regulations that would regulate the works of humans during the scope of employment.
The model justifies financial and economic theories of incentives for the creativity of AI systems and thus enhances and promotes their commercial use.
This solution maintains social and legal stability by not nullifying the applicability of copyright laws by amending and accommodating the laws in a way that suits autonomous systems as well.
Since this model identifies the ownership rights and IP rights of the users, therefore it also automatically imposes accountability on users that would make sure that operations with AI systems are carried on carefully and no infringements occur.
This model provides incentives for every entity that has been involved in the form of licencing to programmers, rights over IP to users, to the government through taxation over the royalties earned, and profits generated through IP, directly from the income of users or from the company’s legal identity, thus not inhibiting the growth of the overall society instead of promoting societal innovation.
Conclusion
The new avatar of robust AI systems has led us to debate the originality of works that have been created by the AI systems, which had been assigned IP rights if they had been created by humans.
In order to promote the incentives to stakeholders associated with AI, the legal system must adopt the flexibility to accommodate the changes to the law and must grant the status to AIs as an artificial person just in a manner a company is considered to be an artificial living legal entity.
The US Constitution under Art. I, § 8, Cl. 8, explicitly states that only humans can be the author; in simple terms, the Constitution has never recognised non-humans as authors. In the case of Naruto v. Slater, the famous monkey selfie case, the district court blatantly refused to recognise the monkey as an author, reasoning that the laws do not extend their application to non-humans [22]. In order to avoid this approach of non-applicability, the laws would have to be modified to be in line with the technologies in place as well as for future technologies. Exploring this issue, the Japanese government in 2016 planned to draft a legal framework in order to protect the works created by AI systems such as music, novels, painting, and so forth [31]. The government agreed to the fact that legislative changes are required to be made in order to protect AI-created works, in order to save them from infringement, and further incentivise the developers to develop the AI systems [32].
With the recognition of AI systems as artificial living entities, the AI system shall also get the status of being authors and inventors, further to which under WMFH doctrine, the work would be transferred to the users [33].
The two-part test that has been suggested eliminates the risk of IP rights being left in ambiguity for AI systems as it shall allow natural persons to gain from the benefits derived. The first part of the test strives to recognise AI’s work that it must be independent and secondly, the AI must have a creative process, rather than just being a mere mechanical machine. Furthermore, by assigning the rights to programmers via explicit and implicit licences and contracts, the overall benefits distributions amongst the stakeholders are equally distributed, thus uplifting and benefitting overall growth and innovation for society as well.
This conclusion acts as a stepping stone here for further significant debates, as once we recognise the creative AIs as authors and thereby transfer the ownership under WMFH doctrine, the further questions that arise here are, how these shall be applicable for the companies working under international space? Would international private and public laws have to step in? How would infringement cases be treated? Will there be any specific ethical guidelines for AI systems? These questions are just a few lying on the tip of the peak right now, and further research shall be required before we put our horses on the cart.
References
1. Ravid, S.Y. and Liu, X., When Artificial Intelligence Systems Produce Inventions: An Alternative Model for Patent Law at The 3A Era.
Cardozo Law Rev.
, 39, 6, 2018.
2. Burrow-Giles Lithographic Co. v. Sarony, 111 U.S. 53 (1883).
3. Turing, A., Computing machinery and intelligence, in:
Computing machinery and intelligence
, Lessing-Druckerei, Mind, Oxford University Press on behalf of the Mind Association, 1950.
4. European Commission, Directorate-General for Research and Innovation, Breque, M., De Nul, L., Petridis, A., Industry 5.0 – Towards a sustainable, human-centric and resilient European industry, Publications Office of the European Union, Luxembourg, 2021.
5. Sejnowski, T.,
The Deep Learning Revolution
, The MIT Press, Cambridge, MA, 2018.
6. Gleeson, N. and Kerrigan, C. (Eds.), Artificial Intelligence Law and Regulation.
Edinburgh Law Rev.
, 27, 1, Edinburgh University Press, Edinburgh, Scotland, 2023.
7. Jones, M.T.,
Artificial intelligence: A systems approach
, Infinity Science Press, Science Press LLC, Hingham, 2008.
8. Russell, S.J.,
Artificial intelligence: A modern approach
(2nd ed.), Prentice Hall/Pearson Education, Upper Saddle River, NJ, 2003.
9. Nilsson, N.J.,
The quest for artificial intelligence: A history of ideas and achievements
, Cambridge University Press, Cambridge, UK, 2010.
10. Abbott, R. and Elgar, E.,
Research handbook on intellectual property and artificial intelligence
, Edward Elgar Publishing, Cheltenham, UK, 2022.
11. Stephens, K., In the age of AI: UK publishes the results of its call for views.
Managing Intell. Prop.
,
https://www.managingip.com/article/2a5cyvdl53v4r4ebzcqgw/in-the-age-of-ai-uk-publishes-the-results-of-its-call-for-views
, 2021.
12. Jordan, M., II and Mitchell, T.M., Machine learning: Trends, perspectives, and prospects.
Sci. (Am. Assoc. Adv. Sci.)
, 349, 6245, 2015.
13. Luger, G.F.,
Artificial intelligence: Structures and strategies for complex problem solving
, Addison-Wesley, Boston, MA, 2005.
14. Scherer, M.U., Regulating artificial intelligence systems: Risks, challenges, competencies, and strategies.
Harv. J. Law Technol.
, 29, 2, 2016.
15. Meshwerks v. Toyota Motor Sales U.S.A., 528 F.3d 1258, 1263 (10th Cir. 2008).
16. Clifford, R.D., Intellectual property in the era of the creative computer program: Will the true creator please stand up?
Tulane Law Rev.
, 71, 6, 1997.
17. Ames, C., Automated Composition in Retrospect: 1956-1986.
Leonardo (Oxford)
, 20, 2, 169–185, 1987.
18. Miller, A., II,
The artist in the machine: The world of AI-powered creativity
, The MIT Press, Cambridge, MA, 2019.
19. Community For Creative Non-Violence v. Reid, 490 U.S. 730, 737 (1989).
20. Aalmuhammed vs. Lee, 202 F.3d 1227, 1234 (9th Cir. 2000).
21. Feist Publications, Inc. v. Rural Telephone Service Co., Inc., 499 U.S. 340, 345 (1991).
22. Naruto v. Slater, 2016 U.S. Dist. Lexis 11041, at 10 (N.D. Cal. Jan. 28, 2016)
23. Wu, A.J., From video games to artificial intelligence: Assigning copyright ownership to works generated by increasingly sophisticated computer programs.
AIPLA Q. J.
,
25
, 1, 131–157, 1997.
24. Final report of the National Commission on new technological uses of copyrighted works.
Computer/Law J.
,
3
, 1, 53–104, 1981.
25. Lee, E., Digital Originality.
Vanderbilt J. Entertain. Technol. Law
, 14, 4, 919–962, 2012.
26. Samuelson., P., Allocating ownership rights in computer-generated works.
Univ. Pittsburgh Law Rev.
, 47, 4, 1185–1228, 1986.
27. Muggleton, S.H. and Chater, N.,
Human-like machine intelligence
, Oxford University Press, Oxford, UK, 2021.
28. Mulgan, G.,
Big mind: How collective intelligence can change our world
, Princeton University Press, Princeton, NJ, 2017.
29. Goldstein vs. California 412 U.S. 546, 561 (1973).
