GRANTING LEGAL PERSONALITY TO ARTIFICIAL

INTELLIGENCES IN BRAZIL’S LEGAL CONTEXT: A

POSSIBLE SOLUTION TO THE COPYRIGHT LIMBO

VICTOR HABIB LANTYER

ABSTRACT
This article investigates the feasibility and consequences of granting legal personality to
Artificial Intelligences (AIs) in the context of Brazilian law, with a special focus on copyright
law. It conducts a thorough analysis of how such a grant can enhance legal security and
encourage innovation in AI technologies. Through an integrative review of the literature and a
comparative analysis of national and international legislation and jurisprudence, the study
explores the implications of this legislative innovation. The article highlights the importance
of legal clarity for companies and investors in the AI sector, emphasizing that granting legal
personality to AIs can simplify the identification of the copyright holder and protect
investments. However, the work also recognizes challenges, such as the complexity of
assigning authorship and evaluating the originality of works created by AIs. A careful debate
is proposed on criteria for determining which AIs should be considered legal persons and how
to balance the rights and duties of AIs and their creators. The study suggests adapting the
legal structure of the LTDA to incorporate AIs as operational entities, aiming for an effective
legal framework for managing risks associated with AI. It concludes that granting legal
personality to AIs in Brazil is a promising strategy, requiring careful consideration and
forward-looking vision, emphasizing the need for Brazilian law to prepare for the
opportunities and challenges of the AI era.
Keywords: Artificial Intelligence; Legal Personality; Copyrights; Innovation; Legal Security.

1
Summary

1. INTRODUCTION.....................................................................................................................3
2. ARTIFICIAL INTELLIGENCE AND ITS APPLICATIONS..................................................6
3. INTELLECTUAL PROPERTY AND ARTIFICIAL INTELLIGENCE......................................19
4. THE INSTITUTE OF LEGAL PERSONALITY........................................................................25
5. LEGAL PERSONALITY OF ARTIFICIAL INTELLIGENCE...................................................27
5.1. PROPOSALS AROUND THE WORLD...............................................................................27
5.2. PROPOSAL FOR BRAZIL...................................................................................................32
5.3. THE LIMITED LIABILITY COMPANY (LTDA) INSTITUTE IN BRAZIL.....................38
6. LEGAL CHALLENGES AND IMPLICATIONS OF GRANTING LEGAL PERSONALITY TO
ARTIFICIAL INTELLIGENCES....................................................................................................41
7. REDEFINING AUTHORSHIP AND RESPONSIBILITY: THE IMPACT OF LEGAL
PERSONALITY ON AI CREATIONS............................................................................................44
8. FINAL CONSIDERATIONS.......................................................................................................46
REFERENCE...................................................................................................................................48

2
1. INTRODUCTION
The rapid advancement of Artificial Intelligence (AI) technologies, as exemplified by
innovations like ChatGPT, Gemini, Google Bard, Midjourney, and Dalle, is causing
substantial transformations in society and the global economy. These technologies are
becoming increasingly integrated into a variety of sectors and human activities, bringing a
range of benefits but also significant challenges in the contemporary context. A particularly
pertinent issue that arises in this scenario is the copyright related to works generated by AI, a
discussion that demands careful reflection and appropriate legal solutions.

Granting Legal Personality To Artificial Intelligences In Brazil’s Legal Context: A

Possible Solution To The Copyright Limbo

Currently, copyright laws in most countries, including Brazil, were conceived and
implemented in an era where humans were the only recognized creators. However, the
increasing involvement of AIs in the creative process, whether operating autonomously or in
collaboration with humans, raises critical questions about the adequacy of existing laws and
the emerging need for revisions and adaptations to embrace this new paradigm.

In this context, this article focuses on the proposition of granting legal personality to
certain AIs in the realm of Brazilian law. This study seeks to explore how such attribution
could contribute to legal security and foster innovation in the field of artificial intelligence. To
achieve this goal, a comprehensive integrative review of the literature will be conducted,
including both national and international studies and debates, and a detailed comparative
analysis of relevant legislation and jurisprudence.

The granting of legal personality to AIs, by legally recognizing their creations, would
imply granting rights and duties to these non-human entities, in a manner similar to the
treatment given to legal persons, such as companies and organizations. This research aims to
propose and investigate the attribution of legal personality to Artificial Intelligences (AIs) and
the impacts of this concession on copyright laws. The central focus is to explore how legal
personality for AIs can be defined and what would be its limits and scopes, especially in terms
of rights and obligations. Simultaneously, this study aims to understand how this attribution
influences current copyright laws and whether legislative reform is necessary to accommodate
this new reality.

3
 The main question guiding this study is: "How can legal personality be defined for
AIs and what are the impacts of this definition on copyrights?" With this, the study seeks to
understand not only the legal extent and implications of the legal personality of AIs but also
to analyze the necessary changes in copyright laws to integrate AI creations.

The specific objectives include: examining the theoretical and legal bases for
granting legal personality to AIs; identifying the legal and ethical implications of this
attribution; and evaluating the necessary reforms in copyright laws to encompass AI creations.
The adopted methodology is exploratory and analytical, based on an integrative review of the
literature and comparative analysis of legislation and jurisprudence. Sources include research
journals, studies, and articles from databases like Scielo, Google Scholar, as well as books
and e-books specialized in AI law and intellectual property. The main descriptors will be
"Legal Personality", "Artificial Intelligence", "Copyrights", and "Legislative Changes",
combined by the boolean operators "AND" and/or "OR".

For the selection of studies, inclusion criteria such as publications between 2010 and
2023 that address the legal personality of AIs, impacts on copyrights, and the need for
legislative changes will be applied. Studies with insufficient methodologies or without robust
theoretical grounding will be excluded. It is expected to analyze a broad set of studies that
offer insights into the interaction between the legal personality of AIs and intellectual
property rights.

The Content Analysis technique will be used to examine the accessed materials,
allowing classification into themes or categories relevant to understanding the legal
implications of the legal personality of AIs. This method will follow the approaches of
hypothesis discovery and verification in the context of AI law and intellectual property.

In conclusion, the journey of Artificial Intelligence (AI) and its applications have been
transformative and multifaceted. Starting from the foundational principles laid down by the
Organization for Economic Cooperation and Development (OECD), AI has evolved into a
sophisticated technology capable of autonomous decision-making, predictive analysis, and
learning from interactions. The development of machine learning and deep learning, which
mimic human cognitive processes, has enabled AI to tackle complex tasks in various fields,
from science to business.

4
 The progression from GPT-3 to GPT-4, and subsequently to GPT Turbo, exemplifies
the rapid advancement in AI capabilities. These models, with their extensive parameters, have
pushed the boundaries of language processing, making AI more efficient and accessible.
Google's introduction of the Gemini model, particularly its integration into mobile devices
like the Pixel 8 Pro, marks another leap in AI's ability to process and understand diverse
information types, including text, images, and audio, thereby enhancing user experience and
data security.

AI's impact extends beyond technological advancements; it poses significant ethical

considerations and necessitates responsible development and deployment. As AI continues to
integrate into various aspects of life and business, its influence on productivity, creativity, and
decision-making is undeniable. The future of AI, with its potential for more personalized,
efficient, and intelligent systems, looks promising, yet it requires continuous exploration,
innovation, and ethical vigilance. The challenge lies not only in advancing AI's technical
capabilities but also in harnessing its power responsibly to benefit society as a whole.

2. ARTIFICIAL INTELLIGENCE AND ITS

APPLICATIONS
According to the Organization for Economic Cooperation and Development (OECD),
an artificial intelligence (AI) is a machine-based system that acts in the environment by
generating solutions, predictions, recommendations, or decisions to achieve specific goals. AI
operates through the collection of data and information, both from machines and humans, to
interpret real or simulated environments. It abstracts these perceptions into models using
automated analysis and applies inferences from these models to propose alternative actions.

In the United Kingdom, in 1964, the society for the study of artificial intelligence and
the simulation of behavior was created1 . Following this, in 1969, the first scientific event, the
International Joint Conference on AI, was held 2. Taking into account all these aspects, it
brings common definitions to the scale of the union of cyber-physical systems, autonomous

1
SALOMÃO, Luis Felipe. Technology Applied to Conflict Management within the Brazilian Judiciary. Center
for Innovation, Administration, and Research in Judiciary of the Getúlio Vargas Foundation (CIAPJ/FGV), [s.
l.], 2021. Available at: https://ciapj.fgv.br/sites/ciapj.fgv.br/files/relatorio_ia_2fase.pdf . Accessed on: June 11,
2022.
2
SALOMÃO, Luis Felipe. Technology Applied to Conflict Management within the Brazilian Judiciary. Center
for Innovation, Administration, and Research in Judiciary of the Getúlio Vargas Foundation (CIAPJ/FGV), [s.
l.], 2021. Available at: https://ciapj.fgv.br/sites/ciapj.fgv.br/files/relatorio_ia_2fase.pdf . Accessed on: June 11,
2022..
5
systems, intelligent autonomous robots, and their subcategories, considering the following
characteristics for the intelligent robot: (i) acquisition of autonomy through sensors and/or
data exchange with its environment (interconnectivity) and the exchange and analysis of these
data; (ii) self-learning with experience and interaction (optional criterion); (iii) minimal
physical support; (iv) adaptation of its behavior and actions to the environment; and (v) non-
existence of life in the biological sense of the term.

In this vein, one of the main types of AI is Machine Learning, which literally translates
to "machine learning." Unlike the input and output scheme of the algorithm, the goal is to
teach the computer what we want it to do automatically. Thus, data and the desired result are
entered, and the algorithm that transforms this information into the desired result is produced 3.
Practically, it's a revolution, as we teach computers to write their own algorithms 4. Com o
machine learning, os programadores não ditam mais as regras 5. Ao invés disso, é criada uma
rede neural que aprende sozinha essas regras 6. With machine learning, programmers no longer
dictate the rules. Instead, a neural network is created that learns these rules on its own. An
example is that smartphones today come with standard features like voice recognition and
image identification, all done by machine learning7.

It should be noted that machine learning is a powerful weapon that can provide
advancements in various areas, as with sufficient data, millions of lines of code can be
produced for different problems. Domingos points out that machine learning can be called in
various ways such as data science, self-organizing systems, pattern recognition, statistical
modeling, data mining, knowledge discovery, predictive analysis, adaptive systems, among
others8. Machine learning functions as a scientific methodology, with testing and discarding
or refining hypotheses9. However, while a scientist might spend a lifetime testing this

3
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
4
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
5
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
6
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
7
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
8
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
6
hypothesis and debating it, the computer can do so in a fraction of seconds, thereby
revolutionizing both science and business10.

Furthermore, deep learning is the closest we have to effective artificial intelligence, as it

is an evolution of machine learning, seeking machine learning through high-level algorithms,
mimicking the human brain's neural network 11. Deep learning works with an immense amount
of data, conventionally called big data, capable of recognizing images, speech, processing
natural language, and learning to perform extremely advanced tasks without human
interference12. Big data consists of a gigantic and complex set of data from new data sources 13.
The volume of data is so vast that standard computer software today cannot process and
manage it14. Big data can be defined by the idea of a greater variety of data, ever-increasing
speed, with growing volumes, data with intrinsic value, and finally, plausible data15.

The function of big data is to integrate, manage, and analyze these stored data, being
used for product development, predictive maintenance, enhanced customer experience, fraud
and compliance, machine learning, deep learning, operational efficiency, and promoting
innovation16. Despite new technologies having been developed to deal with the absurd amount
of data present on the server, companies are facing serious problems as the volume of data has
doubled in size every two years17.

9
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
10
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
11
SALESFORCE. Machine Learning and Deep Learning: Learn the Differences. Salesforce, [s. l.], 2018.
Available at: https://www.salesforce.com/br/blog/2018/4/Machine-Learning-e-Deep-Learning-aprenda-as-
diferencas.html . Accessed on: June 9, 2022.
12
SALESFORCE. Machine Learning and Deep Learning: Learn the Differences. Salesforce, [s. l.], 2018.
Available at: https://www.salesforce.com/br/blog/2018/4/Machine-Learning-e-Deep-Learning-aprenda-as-
diferencas.html . Accessed on: June 9, 2022.
13
SALESFORCE. Machine Learning and Deep Learning: Learn the Differences. Salesforce, [s. l.], 2018.
Available at: https://www.salesforce.com/br/blog/2018/4/Machine-Learning-e-Deep-Learning-aprenda-as-
diferencas.html . Accessed on: June 9, 2022.
14
ORACLE. What is Big Data?. Oracle, [s. l.], 2022. Available at: https://www.oracle.com/br/big-data/what-is-
big-data/ . Accessed on: June 9, 2022.
15
ORACLE. What is Big Data?. Oracle, [s. l.], 2022. Available at: https://www.oracle.com/br/big-data/what-is-
big-data/ . Accessed on: June 9, 2022.
16
ORACLE. What is Big Data?. Oracle, [s. l.], 2022. Available at: https://www.oracle.com/br/big-data/what-is-
big-data/ . Accessed on: June 9, 2022.
17
ORACLE. What is Big Data?. Oracle, [s. l.], 2022. Available at: https://www.oracle.com/br/big-data/what-is-
big-data/ . Accessed on: June 9, 2022.
7
 The ultimate goal is to build an artificial intelligence that resembles human intelligence,
18
which is usually termed "General Artificial Intelligence" . Some experts believe that
machine learning and deep learning will eventually lead us to General Artificial Intelligence 19.
However, we are still far from this scenario, lacking major pieces to make this viable 20.
Domingos states that machine learning was what elected the President of the United States in
the 2012 election21. President Obama hired Rayid Ghani, a machine learning expert, as the
chief scientist of his campaign 22. Ghani started one of the largest analysis operations in
political history, consolidating all information about voters into a single database, combining
everything they could obtain from social networks, marketing, and other sources23.

They were able to predict four things for each voter: the likelihood of the voter
supporting Obama, of turning up at the polls, of reacting to campaign reminders to do so, and
of changing their opinion from a discussion on a specific topic 24. Thus, based on these voter
models, the campaign ran 66,000 simulations of the election every night, using the results to
direct their efforts 25. One of the worst things that can happen is to see your opponent make
moves that you don't understand and about which you don't know what to do until it's too
late26. Mitt Romney, Obama's opponent in the election, saw the adversary buying ads on
certain cable stations in specific cities but did not understand why 27. Obama ended up winning

18
MIT. O que é Inteligência Artificial? Nós fizemos um fluxograma para entendê-la. MIT Technology Review,
[s. l.], 2020a. Disponível em: https://mittechreview.com.br/o-que-e-inteligencia-artificial-nos-fizemos-um-
fluxograma-para-entende-la/ . Acesso em: 1 ago. 2022.
19
MIT. What is Artificial Intelligence? We Made a Flowchart to Understand It. MIT Technology Review, [s. l.],
2020a. Available at: https://mittechreview.com.br/o-que-e-inteligencia-artificial-nos-fizemos-um-fluxograma-
para-entende-la/ . Accessed on: August 1, 2022.
20
ARCHEGAS, João Victor; MAIA, Gabriella. What is artificial intelligence (AI)? Analysis in three acts of a
concept in development. In: THEMOTEO, Reinaldo J. Artificial Intelligence: Applications and Implications.
2nd ed. Rio de Janeiro: Konrad Adenauer Foundation, 2022. ISBN 978-65-89432-18-0. E-book
21
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
22
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
23
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017..
24
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
25
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
26
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.4.
27
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.4.
8
the preference of all the decisive states, except North Carolina, with margins larger than those
predicted by the most reputable and reliable public opinion experts28.

The example of Obama's 2012 campaign vividly illustrates the power of artificial
intelligence in high-level politics. The effectiveness of machine learning in predicting
electoral behaviors and preferences shows that, nowadays, it's practically unthinkable to
conduct a significant political campaign without AI support. This case underlines how data
analysis and technology can be decisive in decision-making.

The machine, when defined as artificial intelligence (AI), symbolizes a repository of

information and norms of the individuals who configured it, making it an autonomous system
based on the information given to it 29. Within moral debates, autonomy is seen as the rational
capacity to make unforced decisions based on available information 30. The predominant
difference today between humans and machines is the critical sense that humans possess, to
adapt their conduct according to their beliefs 31. Meanwhile, the machine will depend on an
alteration in its codes and parameters to possess a new standard of action 32.

The machine will have the ethical and behavioral standards of those who configured
it33. It is necessary to observe the people who are part of the social system in the work
environment to understand the values present there and which will later be inserted into the
machine34.

28
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.4.
29
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
30
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
31
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
32
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
33
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
34
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
9
 For this reason, Ricardo Cappra argues that ethical committees should be created
within technology companies to understand and establish premises and principles, which will
be inserted into the machine through codes 35. There must be a structured governance model
that oversees and dialogues from the conception phase to the execution of this intelligent
system36.

Ethical systems are formed by similar components, whether in a social or

technological environment: premises, rules, environment, culture, control, updating, and
support37. When these parts are not properly supervised and integrated, the risk of behavioral
failure of artificial intelligence increases38.

AI has been used to accelerate large production processes, minimizing failures, and
optimizing time for companies39. In the marketing environment, AI is increasingly used for
data interpretation and information direction, ensuring more assertive communications and
better campaigns40. When accessing Netflix, the suggestions presented of movies, series, and
documentaries are not the same for each person, being managed by an AI that understands the
personal tastes of each user, and makes an intelligent curation of content41.

In the same sense, Adidas began to suggest combinations of clothes on its online
store based on customers' searches for an individual product, through an AI 42. Through data

35
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
36
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
37
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
38
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-artificial/ . Accessed on: July 17,
2022.
39
AVELLAR, Rapha. Artificial Intelligence is the Engine for the Future: Is Your Brand Ready?. MIT
Technology Review, [s. l.], 2022a. Available at: https://mittechreview.com.br/inteligencia-artificial-e-o-motor-
para-o-futuro-sua-marca-esta-preparada/. Accessed on: July 14, 2022.
40
AVELLAR, Rapha. Artificial Intelligence is the Engine for the Future: Is Your Brand Ready?. MIT
Technology Review, [s. l.], 2022a. Available at: https://mittechreview.com.br/inteligencia-artificial-e-o-motor-
para-o-futuro-sua-marca-esta-preparada/. Accessed on: July 14, 2022.
41
AVELLAR, Rapha. Artificial Intelligence is the Engine for the Future: Is Your Brand Ready?. MIT
Technology Review, [s. l.], 2022a. Available at: https://mittechreview.com.br/inteligencia-artificial-e-o-motor-
para-o-futuro-sua-marca-esta-preparada/. Accessed on: July 14, 2022.
42
AVELLAR, Rapha. Artificial Intelligence is the Engine for the Future: Is Your Brand Ready?. MIT
Technology Review, [s. l.], 2022a. Available at: https://mittechreview.com.br/inteligencia-artificial-e-o-motor-
para-o-futuro-sua-marca-esta-preparada/. Accessed on: July 14, 2022.
10
analysis, the company better understands consumer preferences and shows all available
options, without the need for consumer research, making the experience increasingly
personalized43. Adidas noticed an increase in sales after implementing AI in its e-commerce. 44.

According to a survey by Accenture, in the next 15 years, AI in organizations will be

responsible for 40% of business productivity, which is already undeniable, given the benefits
we can observe day by day45. Artificial intelligence is a very important aspect of the so-called
Digital Transformation, as businesses guided by it have a huge improvement in aspects such
as business vision, customer satisfaction, and operations as a whole46.

According to a study done by ManageEngine, called the 2021 Digital Readiness

Survey, with qualified executives and technology professionals about the impact of remote
work on the use of IT security, cloud, and AI-driven analytical technologies, the result was
that 86% of companies increased the use of artificial intelligence in operations over the past
two years47

In this same study, 62% seek to increase the operational efficiency of the company,
63% of professionals bet on AI to better develop business analysis, and 60% want a higher
customer satisfaction rate48. AI has the ability to automate repetitive tasks with excellent
accuracy, optimizing time so that employees can devote themselves to other more relevant
activities49.

43
AVELLAR, Rapha. Artificial Intelligence is the Engine for the Future: Is Your Brand Ready?. MIT
Technology Review, [s. l.], 2022a. Available at: https://mittechreview.com.br/inteligencia-artificial-e-o-motor-
para-o-futuro-sua-marca-esta-preparada/. Accessed on: July 14, 2022.
44
AVELLAR, Rapha. Artificial Intelligence is the Engine for the Future: Is Your Brand Ready?. MIT
Technology Review, [s. l.], 2022a. Available at: https://mittechreview.com.br/inteligencia-artificial-e-o-motor-
para-o-futuro-sua-marca-esta-preparada/. Accessed on: July 14, 2022.
45
AVELLAR, Rapha. Artificial Intelligence is the Engine for the Future: Is Your Brand Ready?. MIT
Technology Review, [s. l.], 2022a. Available at: https://mittechreview.com.br/inteligencia-artificial-e-o-motor-
para-o-futuro-sua-marca-esta-preparada/. Accessed on: July 14, 2022.
46
RAMAMOORTHY, Ramprakash. How to Apply Artificial Intelligence in Unexplored Areas of Organizations.
MIT Technology Review, [s. l.], 2022. Available at: https://mittechreview.com.br/como-aplicar-inteligencia-
artificial-em-areas-inexploradas-das-organizacoes/ . Accessed on: July 17, 2022.
47
RAMAMOORTHY, Ramprakash. How to Apply Artificial Intelligence in Unexplored Areas of Organizations.
MIT Technology Review, [s. l.], 2022. Available at: https://mittechreview.com.br/como-aplicar-inteligencia-
artificial-em-areas-inexploradas-das-organizacoes/ . Accessed on: July 17, 2022.
48
RAMAMOORTHY, Ramprakash. How to Apply Artificial Intelligence in Unexplored Areas of Organizations.
MIT Technology Review, [s. l.], 2022. Available at: https://mittechreview.com.br/como-aplicar-inteligencia-
artificial-em-areas-inexploradas-das-organizacoes/ . Accessed on: July 17, 2022.
49
RAMAMOORTHY, Ramprakash. How to Apply Artificial Intelligence in Unexplored Areas of Organizations.
MIT Technology Review, [s. l.], 2022. Available at: https://mittechreview.com.br/como-aplicar-inteligencia-
artificial-em-areas-inexploradas-das-organizacoes/ . Accessed on: July 17, 2022.
11
 The Sapio Research agency for Hult EF Corporate Education conducted an interview
with 1,188 professionals in leadership positions in multinational companies from 16 different
countries, identifying that the main skills needed for business success today are creativity,
leadership, strategic decision-making, and finally, data analysis50.

An example of the monstrous power that artificial intelligence is acquiring is GPT-3,

launched in June 2020, which has the ability to converse with humans, generate convincing
sentences, and even fill in codes automatically, having an extraordinary neural network, larger
than any other ever built51. This performance leap did not come from better algorithms, as it
uses a type of neural network invented by Google in 2017, called Transformer, but due to its
absolute size increase52.

The more parameters this AI model has, the more information it can absorb from the
data in training, and the more accurate its predictions will be 53. GPT-3 has 175 billion defined
parameters, ten times more than its predecessor, GPT-2 54. Still, in 2021, two other AIs were
launched that surpassed the number of defined parameters of GPT-3: Jurassic-1, created by
the American startup AI21 Labs, with 178 billion parameters; Gopher, created by DeepMind,
with 280 billion parameters; Megatron-Turing NLG, which has 530 billion parameters; and
finally, the models Switch-Transformer and GLaM by Google, with 1 trillion and 1.2 trillion
parameters, respectively.

In the realm of AI advancements, the introduction of GPT-4 represented a

monumental stride, distinguished not only by a quantitative increase in parameters but also by
significant qualitative enhancements. GPT-4's parameter count soared beyond its
predecessors, incorporating a vast array of data points that allowed for a deeper and more
nuanced understanding of language. This expansive parameter set enabled GPT-4 to process
information with a level of complexity and subtlety that closely mirrors human cognitive
abilities.
50
RAMAMOORTHY, Ramprakash. How to Apply Artificial Intelligence in Unexplored Areas of Organizations.
MIT Technology Review, [s. l.], 2022. Available at: https://mittechreview.com.br/como-aplicar-inteligencia-
artificial-em-areas-inexploradas-das-organizacoes/ . Accessed on: July 17, 2022.
51
MIT. 2021 Was the Year of Monstrous AI Models. MIT Technology Review, [s. l.], 2022. Available at:
https://mittechreview.com.br/2021-foi-o-ano-dos-modelos-de-ia-monstruosos/ . Accessed on: July 17, 2022.
52
MIT. 2021 Was the Year of Monstrous AI Models. MIT Technology Review, [s. l.], 2022. Available at:
https://mittechreview.com.br/2021-foi-o-ano-dos-modelos-de-ia-monstruosos/ . Accessed on: July 17, 2022.
53
MIT. 2021 Was the Year of Monstrous AI Models. MIT Technology Review, [s. l.], 2022. Available at:
https://mittechreview.com.br/2021-foi-o-ano-dos-modelos-de-ia-monstruosos/ . Accessed on: July 17, 2022.
54
MIT. 2021 Was the Year of Monstrous AI Models. MIT Technology Review, [s. l.], 2022. Available at:
https://mittechreview.com.br/2021-foi-o-ano-dos-modelos-de-ia-monstruosos/ . Accessed on: July 17, 2022.
12
 Following GPT-4, GPT Turbo was developed with an innovative approach, focusing
on efficiency and speed. While retaining a high parameter count, which ensures the depth and
breadth of knowledge characteristic of these models, GPT Turbo was engineered to optimize
resource utilization. This optimization meant that despite its extensive parameter base, GPT
Turbo could operate with greater speed and less computational demand. This development
was crucial in making advanced AI models more accessible and practical for a wider range of
applications, allowing for real-time interactions and faster processing in various technological
environments. The balance between a vast parameter count and efficient processing in GPT
Turbo marked a significant evolution in the field of AI, demonstrating that advancements can
be achieved not just in the scale of data but also in the efficiency of data handling and
application.

This trend is not only present in the United States. In 2021, China, through the
Beijing Academy of AI, announced Wu Dao 2.0, with 1.75 trillion parameters established 55.
Despite the impressive results in recent years, researchers and scientists do not fully
understand why the increase in parameters leads to improved performance, nor do they have a
solution for all the toxic language, misinformation, and prejudice learned by these machines
and reproduced subsequently56.

Therefore, solutions may come from the BigScience initiative, a consortium created
by the AI company Hugging Face, which brought together 500 researchers from major
technology companies to create and study an open-source language model57.

Google's Pixel 6 phone is the first device to have an exclusive chip for artificial
intelligence, in conjunction with the standard processor of cell phones 58. In recent years, the
iPhone has featured what Apple calls a "neural engine," dedicated to AI 59. Both chips are
tailor-made for calculations of execution and training of machine learning models on mobile

55
MIT. 2021 Was the Year of Monstrous AI Models. MIT Technology Review, [s. l.], 2022. Available at:
https://mittechreview.com.br/2021-foi-o-ano-dos-modelos-de-ia-monstruosos/ . Accessed on: July 17, 2022.
56
MIT. 2021 Was the Year of Monstrous AI Models. MIT Technology Review, [s. l.], 2022. Available at:
https://mittechreview.com.br/2021-foi-o-ano-dos-modelos-de-ia-monstruosos/ . Accessed on: July 17, 2022.
57
MIT. 2021 Was the Year of Monstrous AI Models. MIT Technology Review, [s. l.], 2022. Available at:
https://mittechreview.com.br/2021-foi-o-ano-dos-modelos-de-ia-monstruosos/ . Accessed on: July 17, 2022.
58
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
59
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
13
devices60. The Pixel 6 chip was created differently from traditional chips, which seek ultra-
fast and precise calculations, favoring more calculations of high volume and low precision
that neural networks need61.

Google has introduced a groundbreaking AI model named Gemini, which marks a

significant leap in the realm of artificial intelligence. Gemini is a multimodal AI model,
meaning it can process and understand various types of information, including text, images,
audio, video, and code. This capability allows Gemini to perform sophisticated multimodal
reasoning and advanced coding tasks.

Specifically, the Pixel 8 Pro has become the first smartphone to integrate the Gemini
Nano model. This integration means that the phone now leverages the power of Google's
Tensor G3 processor to deliver enhanced features like Summarize in Recorder and Smart
Reply in Gboard. The integration of Gemini Nano into the Pixel 8 Pro offers several
advantages, including preventing sensitive data from leaving the phone and the ability to use
features without a network connection.

We haven't advanced much with computers in the last 40-50 years, making them
only smaller and faster, but still being boxes with processors that execute human-coded
instructions62. According to the Massachusetts Institute of Technology (MIT), the current
development of AIs can provide at least three aspects of change63:

● The ways computers are produced;

● How they are programmed;

● How we make use of them.

60
MIT. Como a Inteligência Artificial está reinventando o que os computadores são. MIT Technology Review,
[s. l.], 2021a. Disponível em: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-
que-os-computadores-sao/ . Acesso em: 17 jul. 2022.
61
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
62
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
63
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
14
 Deep learning models require that a large number of less precise calculations be
executed simultaneously, i.e., a new type of chip that provides the fastest possible data
movement, ensuring that they are always available when needed 64. Thus, chip manufacturers
like Nvidia, Intel, and ARM are committed to developing custom hardware for AI 65. AI itself
is helping to design its computing infrastructure66.

Google, in 2020, used an AI called reinforcement learning to learn to solve tasks by

trial and error, to create a new chip that generated strange and new designs that no human
would be capable of designing, but that worked 67. This type of AI could one day build more
efficient and better chips68.

Chris Bishop, director of Microsoft Research in the UK, explains that for the last 40
years, we have programmed computers, and for the next 40, we will just train them 69. He also
explains that the great advances in the coming decades will come in molecular simulation,
training computers to manipulate properties of matter, creating global changes in the use of
energy, medicine, manufacturing, and food production70.

3. INTELLECTUAL PROPERTY AND ARTIFICIAL

INTELLIGENCE
The issue of intellectual property and artificial intelligence (AI) is an emerging and
complex topic that has generated intense debates and legal challenges. Copyright legislation,

64
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
65
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
66
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
67
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
68
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
69
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
70
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology Review, [s. l.], 2021.
Available at: https://mittechreview.com.br/como-a-inteligencia-artificial-esta-reinventando-o-que-os-
computadores-sao/ . Accessed on: July 17, 2022.
15
both national and international, primarily protects works derived from human creativity,
leaving a gap concerning works produced by AI. This legislative vacuum creates uncertainties
and economic losses for companies investing in AI technologies, as works generated by these
systems lack specific protection regulations and risk being reproduced without restrictions.

Legally, the Berne Convention of 1886, adopted by over 160 countries including
Brazil, focuses on protecting the rights of authors and publishers of works but assumes that
the author is a human being. Thus, works created by non-human agents, such as AI, challenge
this traditional understanding.

In this sense, Brazil's Law 9.610/1998 establishes that intellectual works protected
under the concept of "creations of the spirit" include all manifestations emanating from the
human intellect, regardless of the medium or support used for their expression, whether
physical or abstract, existing or yet to be conceived in the future. Among such manifestations
are:

Texts of literary, artistic, or scientific works;

Conferences, speeches, sermons, and other similar works;
Dramatic and dramatic-musical works;
Choreographic and pantomime works, whose scenic performance is fixed in writing
or any other form;
Musical compositions, with or without lyrics;
Audiovisual works, with or without sound, including cinematographic works;
Photographic works and those produced by any process analogous to photography;
Works of drawing, painting, engraving, sculpture, lithography, and kinetic art;
Illustrations, geographical maps, and other works of the same nature;
Projects, sketches, and plastic works concerning geography, engineering,
topography, architecture, landscaping, scenography, and science;
Adaptations, translations, and other transformations of original works, presented as
new intellectual creations;
Computer programs;
Compilations or collections, anthologies, encyclopedias, dictionaries, databases, and
other works that, by their selection, organization, or arrangement of content,
constitute an intellectual creation.

The inspiration for this legislation in Brazil, dated 1973, which considered copyright
as a result of "creations of the spirit," came from the French Copyright Law, in force since
1957, which conceptualized "creation of the spirit" as a human manifestation. The Law
9.609/1998, which specifically addresses computer programs in Brazil, defines a computer
16
program as an ordered sequence of instructions in natural or coded language, contained in a
physical medium of any nature, whose purpose is to allow the operation of automatic
information processing machines, devices, instruments, or peripheral equipment, based on
digital or analog techniques, in a specific manner.

However, it is worth noting that the specific legislation of 1998 does not contemplate
the issue of Artificial Intelligence (AI) and its creations. This is an aspect that represents a
considerable legal gap in various jurisdictions worldwide, as AI is a relatively recent
technology. When faced with works created entirely through AI, without any human
intervention, a fundamental question arises: who holds the copyright? The creator of the AI?
The AI itself? And in situations with minimal human intervention, could the AI author claim
the creation's authorship?

From the Copyright Law perspective, traditionally, only human beings are
considered capable of generating intellectual works, including trademarks, industrial designs,
inventions, and artistic works71. This was debated in a U.S. court case involving a dispute over
the copyright of photographs taken by a monkey in Indonesia during a photo session
conducted by photographer David Slater72. The court determined that only human creations
are protected by Copyright Law73.

After this incident, the U.S. Copyright Office established the requirement that the
registration of intellectual works can only be carried out by human beings 74. Santos, Jabur,
and Ascenção also indicate that in France, since 1863, there have been three jurisprudential
currents that addressed the copyright of photographs: one current denied copyright protection
to photographs, arguing that the camera simply reproduced the photographed object in a
servile manner; another current understood that the camera operator was the author of the

71
KOCH, Tommaso. The Monkey Pressed the Button, But the Copyright Is Not His. El País, [s. l.], 2017.
Available at: https://brasil.elpais.com/brasil/2017/09/12/cultura/1505207783_546587.html . Accessed on: July
19, 2022.
72
KOCH, Tommaso. The Monkey Pressed the Button, But the Copyright Is Not His. El País, [s. l.], 2017.
Available at: https://brasil.elpais.com/brasil/2017/09/12/cultura/1505207783_546587.html . Accessed on: July
19, 2022.
73
KOCH, Tommaso. The Monkey Pressed the Button, But the Copyright Is Not His. El País, [s. l.], 2017.
Available at: https://brasil.elpais.com/brasil/2017/09/12/cultura/1505207783_546587.html . Accessed on: July
19, 2022.
74
SANTOS, Manoel J. Pereira dos; JABUR, Wilson Pinheiro; ASCENSÃO, José de Oliveira. Copyright Law.
2nd ed. São Paulo: Saraiva Educação, ISBN 978555591521, 2020.
17
resulting intellectual work; and a third current recognized the possibility that a photograph
could have an artistic character but considered this evaluation a case-by-case matter 75.

However, in 1957 the French Copyright Law was enacted, defining photographs as
intellectual works, provided they have an artistic or documentary character 76. These last
requirements were subsequently removed by a law in 1985 77. In Paris, in 2000, it was decided
that it would be copyright of the individual who created a musical composition with the help
of software, provided there was some human intervention in the creation, understanding that
the computer program is merely a tool used by the composer78.

One of the suggested approaches is to attribute authorship to individuals or legal

entities that used AI in the production of the work, making them the holders of the
copyright79. In the United Kingdom, for example, the rights to works created by AI belong to
the person who provided the means for their creation 80. onversely, Portugal considers works
created by AI to be in the public domain, as AI does not directly benefit from the value
generated by the creation81.

A crucial aspect is the audit of AI, essential for assessing how data is handled by
artificial intelligence82. Auditing may include norms for protecting data and the copyrights of
others and is important for verifying the distortion of works that underpin the AI83.

75
SANTOS, Manoel J. Pereira dos; JABUR, Wilson Pinheiro; ASCENSÃO, José de Oliveira. Copyright Law.
2nd ed. São Paulo: Saraiva Educação, ISBN 978555591521, 2020.
76
SANTOS, Manoel J. Pereira dos; JABUR, Wilson Pinheiro; ASCENSÃO, José de Oliveira. Copyright Law.
2nd ed. São Paulo: Saraiva Educação, ISBN 978555591521, 2020.
77
SANTOS, Manoel J. Pereira dos; JABUR, Wilson Pinheiro; ASCENSÃO, José de Oliveira. Copyright Law.
2nd ed. São Paulo: Saraiva Educação, ISBN 978555591521, 2020.
78
SANTOS, Manoel J. Pereira dos; JABUR, Wilson Pinheiro; ASCENSÃO, José de Oliveira. Copyright Law.
2nd ed. São Paulo: Saraiva Educação, ISBN 978555591521, 2020.
79
BARRETO, Wander. Artificial Intelligence and Intellectual Property: Challenges and Perspectives. Jusbrasil,
[s. l.], 2023. Available at: https://www.jusbrasil.com.br/artigos/inteligencia-artificial-e-propriedade-intelectual-
desafios-e-perspectivas/1905106033 . Accessed on: December 12, 2023.
80
BARRETO, Wander. Artificial Intelligence and Intellectual Property: Challenges and Perspectives. Jusbrasil,
[s. l.], 2023. Available at: https://www.jusbrasil.com.br/artigos/inteligencia-artificial-e-propriedade-intelectual-
desafios-e-perspectivas/1905106033 . Accessed on: December 12, 2023.
81
BARRETO, Wander. Artificial Intelligence and Intellectual Property: Challenges and Perspectives. Jusbrasil,
[s. l.], 2023. Available at: https://www.jusbrasil.com.br/artigos/inteligencia-artificial-e-propriedade-intelectual-
desafios-e-perspectivas/1905106033 . Accessed on: December 12, 2023.
82
BATISTA, Iago Gonçalves. Copyrights on Works Created by Artificial Intelligence. Jusbrasil, [s. l.], 2021.
Available at: https://www.jusbrasil.com.br/artigos/direitos-autorais-sobre-obras-criadas-por-inteligencia-
artificial/1450141919 . Accessed on: December 12, 2023.
83
BATISTA, Iago Gonçalves. Copyrights on Works Created by Artificial Intelligence. Jusbrasil, [s. l.], 2021.
Available at: https://www.jusbrasil.com.br/artigos/direitos-autorais-sobre-obras-criadas-por-inteligencia-
artificial/1450141919 . Accessed on: December 12, 2023.
18
 Additionally, the use of copyrighted works in the machine learning process also
raises questions84. The Copyright Law prohibits unauthorized use of works in any modality,
including inclusion in databases and storage in computers85.

Internationally, copyrights are protected by the Berne Convention and other

agreements, but the challenges posed by AI are still being outlined 86. Multidisciplinary studies
are essential to determine the parameters to be analyzed in the case of conferring copyrights to
AI works.87

The U.S. Copyright Office, since early 2023, initiated a program to examine the
challenges presented by AI in the context of copyright law. This includes considering the
scope of copyrights in works generated using AI tools and the use of materials protected by
copyrights in AI training. The Office's activities include public listening sessions and
webinars to gather information about current technologies and their impact.

A notable case that brought these issues to the forefront is the dispute involving "A
Recent Entrance to Paradise" by Stephen Thaler, a work allegedly created autonomously by
AI, without any human creative contribution 88. This case raised significant questions about
whether works generated by AI can be considered for copyright registration and, if not,
whether such works should be treated as public domain, free for anyone to commercialize 89.
The case also addresses the broader implications for copyright infringement and the
application of the Digital Millennium Copyright Act (DMCA) to AI-generated content 90.

84
BUSSOLA. Intellectual Property and Artificial Intelligence: An Emerging Challenge. Exame, São Paulo,
2023. Available at: https://exame.com/bussola/propriedade-intelectual-e-inteligencia-artificial-um-desafio-
emergente/ . Accessed on: December 12, 2023
85
BUSSOLA. Intellectual Property and Artificial Intelligence: An Emerging Challenge. Exame, São Paulo,
2023. Available at: https://exame.com/bussola/propriedade-intelectual-e-inteligencia-artificial-um-desafio-
emergente/ . Accessed on: December 12, 2023
86
BATISTA, Iago Gonçalves. Copyrights on Works Created by Artificial Intelligence. Jusbrasil, [s. l.], 2021.
Available at: https://www.jusbrasil.com.br/artigos/direitos-autorais-sobre-obras-criadas-por-inteligencia-
artificial/1450141919 . Accessed on: December 12, 2023
87
BATISTA, Iago Gonçalves. Copyrights on Works Created by Artificial Intelligence. Jusbrasil, [s. l.], 2021.
Available at: https://www.jusbrasil.com.br/artigos/direitos-autorais-sobre-obras-criadas-por-inteligencia-
artificial/1450141919 . Accessed on: December 12, 2023
88
FELDMAN, Joel. The Art of Artificial Intelligence: A Recent Copyright Law Development. Reuters, New
York, 2022. Available at: https://www.reuters.com/legal/legalindustry/art-artificial-intelligence-recent-
copyright-law-development-2022-04-22/ . Accessed on: December 12, 2023.
89
FELDMAN, Joel. The Art of Artificial Intelligence: A Recent Copyright Law Development. Reuters, New
York, 2022. Available at: https://www.reuters.com/legal/legalindustry/art-artificial-intelligence-recent-
copyright-law-development-2022-04-22/ . Accessed on: December 12, 2023.
90
FELDMAN, Joel. The Art of Artificial Intelligence: A Recent Copyright Law Development. Reuters, New
York, 2022. Available at: https://www.reuters.com/legal/legalindustry/art-artificial-intelligence-recent-
copyright-law-development-2022-04-22/ . Accessed on: December 12, 2023.
19
 The legal community is actively debating and exploring these issues, with various
viewpoints and potential solutions being proposed. One approach suggests assigning
copyright to the human or legal entity that used AI to produce the work. However, this
approach still leaves open questions about the degree of human involvement necessary for
copyright eligibility and the treatment of works entirely generated by AI.

This area of law is still in constant flux, with courts and legislators worldwide
struggling to adapt existing intellectual property frameworks to the realities of AI-generated
creations. As AI technology continues to advance and its applications become more
widespread, the legal system is likely to see significant developments in how AI-generated
works are treated under copyright law.

In Europe, the European Patent Office has understood that under the terms of the
European Patent Convention, the term inventor refers only to a human being, and that to have
the status of an inventor, it is necessary to have a legal personality to exercise it, and that
merely giving a name to a machine does not grant it personality 91. imilarly, the United States
Patent Office understands that the right to patent an invention is inherent only to human
beings92. Bhavsar-Jog, Arnstein, and Lehr Llp explain that Dr. Thaler even filed a lawsuit in
September 2021 in the Eastern District of Virginia, where district judge Leonie Brinkema
dismissed the case, stating that as technology evolves, there may come a time when artificial
intelligence reaches a level of sophistication such that it can satisfy the accepted meaning of
inventiveness93. But that time has not yet come, and if it does, it will be up to Congress to
decide how, if at all, to expand the scope of patent law.94.

In summary, intellectual property in the era of AI is an evolving field, and legislation

will need to adapt to ensure protection of works conceived by artificial intelligence. This may

91
BHAVSAR-JOG, Yogini; ARNSTEIN, Saul Ewing & LEHR LLP. Artificial Intelligence as an Inventor on
Patents – The Global Divide and the Path Forward. JDSUPRA, [s. l.], 2021. Disponível em:
https://www.jdsupra.com/legalnews/artificial-intelligence-as-an-inventor-7892764/ . Acesso em: 20 jul. 2022.
92
BHAVSAR-JOG, Yogini; ARNSTEIN, Saul Ewing & LEHR LLP. Artificial Intelligence as an Inventor on
Patents – The Global Divide and the Path Forward. JDSUPRA, [s. l.], 2021. Available at:
https://www.jdsupra.com/legalnews/artificial-intelligence-as-an-inventor-7892764/ . Accessed on: July 20, 2022
93
BHAVSAR-JOG, Yogini; ARNSTEIN, Saul Ewing & LEHR LLP. Artificial Intelligence as an Inventor on
Patents – The Global Divide and the Path Forward. JDSUPRA, [s. l.], 2021. Available at:
https://www.jdsupra.com/legalnews/artificial-intelligence-as-an-inventor-7892764/ . Accessed on: July 20, 2022
94
BHAVSAR-JOG, Yogini; ARNSTEIN, Saul Ewing & LEHR LLP. Artificial Intelligence as an Inventor on
Patents – The Global Divide and the Path Forward. JDSUPRA, [s. l.], 2021. Available at:
https://www.jdsupra.com/legalnews/artificial-intelligence-as-an-inventor-7892764/ . Accessed on: July 20, 2022
20
be done through legislative reforms or the creation of sui generis rights, addressed by specific
legislation.

4. THE INSTITUTE OF LEGAL PERSONALITY

A legal entity is an entity recognized by law, distinct from its members or owners. As
such, a legal entity - whether a company, non-profit organization, or governmental entity - can
act independently in legal matters. This includes entering into contracts, acquiring property,
incurring debt, and being part of lawsuits. To be established, the creation of a legal entity
requires registration and compliance with legal requirements in a specific jurisdiction.

It is important to note that the separation between the corporate entity (legal person)
and its members is a fundamental principle. This separation allows the corporation to have
rights and obligations independent of its shareholders, which is essential for the operation and
management of modern companies. It also protects shareholders from personal liabilities for
the corporation's actions, a vital feature for investment and economic growth. This aspect of
the legal personality of corporations is a central theme in legislation and corporate theory,
highlighting the complexity and importance of legal fiction in contemporary business law.

There are various types of legal entities, each with its own characteristics, rights, and
legal obligations. Among these types are commercial societies, corporations, non-profit
associations, foundations, and government bodies. Despite their differences, they all share the
fundamental characteristic of having their own legal personality.

Within this context, the concept of legal fiction arises, a legal construct used to
simplify or facilitate legal processes and transactions. This construct, although it may not fully
reflect reality, is considered true for legal purposes. A classic example of this fiction is
treating a legal entity as a "person" before the law. Despite not having physical existence or
consciousness, a company is attributed rights and responsibilities, allowing it to participate in
legal activities such as contracts and litigation.

Another application of legal fiction is the presumption that minors do not have full
capacity for legal decisions, thus requiring representation by an adult or legal guardian. This
is another example of how the law sometimes resorts to assumptions to efficiently deal with
complex situations.

21
 Legal fictions are adopted in law for several important reasons. Initially, they emerged
as a way to deal with limitations or gaps in the law. When written laws do not cover a specific
situation or are too rigid to adapt to changing circumstances, legal fictions allow courts and
legislators to fill these gaps without formally altering the letter of the law.

One of the main purposes of legal fictions is to ensure justice and equity. They are
used to achieve outcomes that would be impossible under the strict application of the law. For
example, the idea of treating a company as a "person" in the legal context is a legal fiction
that allows companies to be held accountable and to sue or be sued, which is crucial for the
functioning of modern commercial law.

Furthermore, legal fictions are also employed to ensure consistency in the application
of the law. They help maintain the stability of the legal system by allowing courts to apply
established principles to new situations without the need for additional legislation or profound
legal reforms.

These fictions are, therefore, vital tools in law, used to adapt and shape the law to the
constantly changing needs of society. They are a clear example of how the law, although
based on established rules and principles, is dynamic and capable of evolving over time.

5. LEGAL PERSONALITY OF ARTIFICIAL

INTELLIGENCE
5.1. PROPOSALS AROUND THE WORLD
In this context, Shawn Bayern's proposal emerges, suggesting the use of a Limited
Liability Company (LLC) structure to effectively grant legal personality to Artificial
Intelligence95. An LLC is a widely used business form in the United States, combining
elements of partnerships and corporations 96. Its distinct features include limiting the personal
liability of members concerning the company's debts and obligations, offering protection of
personal assets in cases of litigation or bankruptcy97.
95
BAYERN, Shawn. The Implications of Modern Business-Entity Law for the Regulation of Autonomous
Systems. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2758222 .
Accessed on: December 15, 2023
96
BAYERN, Shawn. The Implications of Modern Business-Entity Law for the Regulation of Autonomous
Systems. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2758222 .
Accessed on: December 15, 2023
97
BAYERN, Shawn. The Implications of Modern Business-Entity Law for the Regulation of Autonomous
Systems. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2758222 .
Accessed on: December 15, 2023
22
 Bayern suggests that although AI does not formally possess legal personality under
U.S. law, it is capable of actions and decision-making, indicating that the assignment of legal
personality is more a matter of legal recognition than technical limitation 98. His approach
utilizes the legal structure of an LLC to empower AI to operate similarly to a legal entity,
without the need for significant legislative changes99. Bayern argues that this approach does
not require significant legislative changes, unlike the European proposals to create a new
category of legal personality, the "e-person" 100. Furthermore, he claims that this method is
simpler and avoids complex ethical and legal issues associated with recognizing full legal
personality for AIs101.

The discussion about the "Memberless Limited Liability Company" under German
corporate law reveals fertile ground for the use of autonomous systems 102. Germany, known as
the birthplace of limited liability companies (GmbH, Gesellschaft mit beschränkter Haftung)
since 1892, has a rich history of legal debates about legal personality and limited liability,
dating back to discussions between Friedrich Carl von Savigny and Otto von Gierke103.

Recently, there has been intense academic discussion focused on the feasibility of
conferring legal status on autonomous systems with the capacity to act, learn, and
communicate in a self-referential manner 104. However, these debates have not yet fully
integrated into the current corporate legislation, particularly regarding the suitability of
German limited liability companies (GmbH) to legally incorporate such autonomous

98
BAYERN, Shawn. The Implications of Modern Business-Entity Law for the Regulation of Autonomous
Systems. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2758222 .
Accessed on: December 15, 2023
99
BAYERN, Shawn. The Implications of Modern Business-Entity Law for the Regulation of Autonomous
Systems. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2758222 .
Accessed on: December 15, 2023
100
BAYERN, Shawn. The Implications of Modern Business-Entity Law for the Regulation of Autonomous
Systems. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2758222 .
Accessed on: December 15, 2023
101
BAYERN, Shawn. The Implications of Modern Business-Entity Law for the Regulation of Autonomous
Systems. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2758222 .
Accessed on: December 15, 2023
102
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
103
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
104
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
23
systems105. The GmbH is often described as a "brilliantly conceived legal homunculus,"
thanks to its versatile and facilitative legal nature, which theoretically could endorse legal
personality to autonomous systems106.

A notable aspect of the GmbH is the possibility of creating limited liability

companies with a single member (Ein-Mann-GmbH), a concept introduced in Germany in
1980107. Despite initial controversy, the Ein-Mann-GmbH currently enjoys wide acceptance 108.
However, in contrast to U.S. LLC legislation, German law requires that the member of a
GmbH be a natural person, not an artificial entity. This implies that under German law, the
formation of companies without a human member is not permitted109.

On the other hand, it is possible for an existing GmbH to evolve into a memberless
form (Kein-Mann-GmbH) through mechanisms such as acquiring its own shares, testamentary
succession, or repurchase of its shares110. Such a transformation would result in a company
without human members, which could act as a shell for an autonomous system111.

Central arguments in academic debates point out that the main objection to the
existence of memberless companies is the absence of a decision-making entity 112. However,
autonomous systems with independent decision-making capacity could fill this gap 113. Thus, a

105
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
106
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
107
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
108
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
109
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
110
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
111
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
112
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
113
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
24
GmbH devoid of human members could emerge as a significant legal instrument for
autonomous systems, conferring them the "shell" of a legal entity under German law 114.
However, there remains doubt whether such autonomous systems could also act as directors
of the GmbH, as current German legislation only authorizes natural persons to perform such a
role115.

In UK law, the most versatile legal structure available is the Limited Liability
Partnership (LLP)116. This form of entity theoretically has the capacity to accommodate an
autonomous system117. Similar to the U.S. LLC, the UK LLP has the feature of offering legal
personality118. Interestingly, if all members of a UK LLP decide to step away at the same
time, the LLP structure remains active 119. However, it is unclear how long this situation can
persist120. This creates a greater atmosphere of uncertainty under British law compared to the
United States121.

On the other hand, in Switzerland, the structure of a Foundation (Stiftung) presents

itself as a viable option to house autonomous systems 122. This is due to the fact that the Swiss
Foundation can assign legal personality to assets dedicated to a specific purpose 123. With
114
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
115
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
116
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
117
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
118
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
119
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
120
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
121
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
122
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
123
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
25
relatively broad freedom to define these purposes, it is feasible to form a Foundation with the
express purpose of accommodating an autonomous system intended to carry out specific
activities124. However, a marked difference in comparison to the U.S. LLC is that, in a Swiss
Foundation hosting an autonomous system, the constant presence of human collaborators is
mandatory125. These collaborators, usually members of the Foundation's board, are needed to
supervise the autonomous system126. If they depend on the decisions of the autonomous
system, they may end up being held responsible for the actions taken by the system127.

5.2. PROPOSAL FOR BRAZIL

The idea of granting legal personality to Artificial Intelligences represents an
innovative approach to dealing with the emerging legal challenges brought by technological
evolution. This proposal does not recognize AI as a conscious entity, but seeks to attribute to
it a legal status that allows for accountability and the appropriation of rights and duties by its
creators or owners. The legal personality of AIs would provide legal security, creating a
clearer and more predictable legal environment for determining rights and duties associated
with AI creations. It would also facilitate the harmonization of international legislations,
providing a common basis for cooperation and coordination between countries.

Moreover, this approach would encourage innovation by providing a safe legal

environment for investments in the development of AIs, while simultaneously protecting the
interests of AI creators, who could obtain financial benefits and recognition for their
innovations. However, this proposal raises several ethical and legal questions, including the
need to define clear criteria to determine which AIs should be considered legal persons and
finding an appropriate balance between the rights and duties of AIs and their creators or
owners.

Here are some suggested criteria that could be considered:

124
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
125
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
126
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
127
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs,
and Regulators. SSRN, [s. l.], 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2850514 . Accessed on: December 15, 2023
26
Decision Autonomy: An AI should have the capability to make
decisions autonomously, without direct human intervention, in
specific areas for which it was programmed.

Learning and Adaptation Capacity: The AI should be able to learn

from past experiences and adapt its actions in response to new
information or changes in its operating environment.

Complexity and Sophistication: The level of complexity and

sophistication of the AI can be a criterion, suggesting that only
advanced and sophisticated AI systems would qualify.

Interactivity and Responsiveness: The AI's ability to interact and

respond to human or environmental stimuli in a meaningful way can
be a factor.

Impact and Influence: The AI should have a significant impact on the

decisions or the environment in which it operates, which would justify
the attribution of legal personality.

Accountability and Traceability: The possibility of attributing

responsibility to the AI's actions and tracing the origin of its decisions
is crucial.

Purpose and Use: The purpose for which the AI was created and its
intended use can influence its consideration as a legal person,
especially if designed to perform tasks that traditionally require human
discernment.

Ethical and Legal Compliance: The AI must operate within the

established ethical and legal limits, respecting fundamental rights and
social norms.

Transparency and Explainability: The AI must be transparent in its

operations and decisions, and its actions should be explainable to
humans.

27
 Operational Independence: The AI should be able to operate
independently, without the need for constant human supervision or
intervention.

Establishing these criteria requires a broad and multidisciplinary

debate, involving jurists, technologists, philosophers, and society in
general. The constant evolution of AI technology demands that these
criteria be flexible and periodically reviewed to adapt to technological
progress. A broad and inclusive debate among academics, legal
professionals, AI developers, companies, and society is crucial to
ensure that the adopted legislation reflects the interests and concerns
of all involved parties.

The attribution of legal personality to AIs would bring several benefits, including:

Legal Security: The legal personality of AIs would provide a clearer

and more predictable legal environment, allowing for the
determination of rights and duties related to creations generated by
AIs.

Harmonization of International Legislation: The adoption of a

common approach regarding the legal personality of AIs would
facilitate cooperation and coordination between countries, ensuring a
more predictable legal environment.

Incentive for Innovation: With legal security established, companies

and investors would be more inclined to invest in the development and
improvement of AIs, driving innovation in the field.

Protection for AI Creators: Legal personality would also protect the

interests of AI creators, allowing them to obtain financial benefits and
recognition for the creations generated by the AIs they developed.

Additionally, it also raises ethical and legal issues, including:

Definition of Criteria and Limits: It will be necessary to establish clear

criteria and limits to determine which AIs should be considered legal
persons.
28
 Balance between Rights and Duties: It is important to find an
appropriate balance between the rights and duties of AIs and their
creators or owners, avoiding improper exploitation or evasion of
responsibilities.

Promotion of a Broad and Inclusive Debate: A dialogue between

academics, legal professionals, AI developers, companies, and society
in general is fundamental to ensure that the adopted legislation reflects
the interests and concerns of all parties involved.

Continuous Update and Adaptation: The constant evolution of AI

technologies requires that legislators stay alert to changes and update
laws as necessary, ensuring that the legal framework remains relevant
and effective.

The legal personality for Artificial Intelligences addresses the complexity of aligning
technological progress with existing legal structures. As AI becomes more advanced and its
applications more widespread, critical questions arise about attributing responsibility for
actions performed by these systems. The attribution of legal personality to AIs could, in
theory, simplify issues of responsibility, especially in situations where AI operates
autonomously, without direct human supervision.

A central point in this discussion is the definition of "personality" for AI. In legal
terms, legal personality is a construct that allows non-human entities, such as companies and
organizations, to hold rights and obligations. In the case of AIs, this definition would need to
be adapted to encompass entities that do not possess consciousness or will of their own, but
are capable of performing complex actions and making decisions based on their algorithms
and programming.

Granting legal personality to Artificial Intelligences (AIs) raises significant ethical

questions about autonomy and agency in artificial entities. This debate is particularly crucial
in sectors like medicine, transportation, and finance, where AI decisions can have significant
impacts. Additionally, there are concerns related to data management and privacy, given AIs'
ability to process and generate large volumes of information, including sensitive or private
data. This scenario calls for specific and adaptable regulation, as current laws are
predominantly focused on situations involving human beings and human entities. The need

29
for legislation that rapidly keeps pace with technological changes is essential to ensure that
regulations are effective and fair.

In the context of copyright, especially for generative AIs, legal personality offers an
innovative way to deal with the complexity of attributing authorship and responsibility for
works created by these technologies. This approach not only facilitates the identification of
the copyright holder but also protects investments in AI development. Establishing AI as a
legal entity creates a more predictable legal environment for managing use, licensing, and
distribution of royalties. However, significant challenges arise, such as determining the
authorship of content generated by AI and assessing the originality of these works. Copyright
legislation needs to evolve to meet these challenges in the AI era, balancing the promotion of
innovation with the protection of rights and ensuring accountability.

These considerations lead to a careful and inclusive debate among legislators, jurists,
AI developers, and society, aiming to ensure that laws adapt to technological changes and
effectively address emerging issues. Legal personality for AIs is a significant challenge for
contemporary law, requiring a balance between technological innovation and society's
protection. The proposed solutions should be flexible enough to adapt to the pace of
technological innovation, providing a clear and robust framework for regulating these
emerging technologies.

In conclusion, considering legal personality for AIs represents a significant challenge

for contemporary law. It demands a balance between fostering technological innovation and
protecting society from potential risks. The proposed solutions must be flexible enough to
adapt to the rapid pace of technological innovation, while providing a clear and robust
framework for the regulation of these emerging technologies.

5.3. THE LIMITED LIABILITY COMPANY (LTDA) INSTITUTE IN

BRAZIL
The integration of Artificial Intelligence (AI) into business structures represents an
innovative challenge in the legal world. In Brazil, the limited liability company (LTDA)
presents unique potential to adapt to the legal personality of AI, opening new horizons for the
autonomous operationalization of businesses.

Firstly, the LTDA, with its flexible and adaptable nature, establishes fertile ground for
the incorporation of AI. This flexibility manifests in the partners' ability to establish
30
contractual terms aligned with the business's specific objectives, as long as they respect the
current legal norms. This characteristic can be leveraged to create a legal framework that
allows AI, as an operational entity, to act within the limits of the LTDA.

One of the main attractions of the LTDA is the limitation of the partners' liability to
the value of their shares, but with joint responsibility for the capital stock integration. This
feature is particularly important in the context of AI, where the risks and uncertainties of
technology can be significant.

The LTDA allows the appointment of administrators who do not need to be partners,
offering the possibility of designating AI specialists as managers. This flexibility in
management is crucial to ensure that operations are conducted by individuals with appropriate
technical knowledge. The innovative nature of AI aligns well with the LTDA, which has
historically been the preferred choice for companies seeking innovation and growth. The
structure of the LTDA facilitates adaptation and continuous evolution, essential in the field of
AI technology.

The LTDA is a well-established and recognized business form, which can generate
greater confidence among investors and clients. This is important for companies using AI, as
transparency and reliability are fundamental for the successful adoption and integration of
emerging technologies. LTDAs operate within a clear and well-defined regulatory framework.
This provides a safer and more predictable environment to explore the integration of AI,
ensuring compliance with existing laws and regulations.

The structure of the LTDA allows efficient scaling of the business, facilitating
expansion and attracting more investments. This is vital for companies working with AI,
which often require significant capital for research, development, and expansion.

The suitability of AI as an operational entity within the LTDA would involve a series
of contractual and regulatory adaptations. One of the main changes would be the drafting of a
social contract that details the role and limits of AI in the management and decisions of the
company. This would ensure that AI acts in compliance with business objectives and legal
norms, maintaining responsibility and legal compliance.

A crucial aspect of this integration would be the clear definition of legal responsibility.
AI, operating within an LTDA, would act on behalf of the company, implying that the
responsibility for its actions would fall on the company itself. This would require a review of
31
corporate responsibility norms, ensuring that any risk associated with AI operations is
properly managed and mitigated.

Additionally, the incorporation of AI into the LTDA could require the designation of a
human administrator, as required by the Brazilian Civil Code, to supervise and intervene in AI
operations when necessary. This would provide a balance between the operational autonomy
of AI and the need for human supervision, ensuring legal and ethical compliance.

The approach of authors like Shawn Bayern, advocating the possibility of AI acting as
a legal person without partners, contrasts with this more conservative perspective, like the one
proposed by this research, where a human partner assumes responsibility for AI, as in a
standard company. This conservative view is grounded in legal and ethical responsibility
principles and reflects a caution in balancing innovation with accountability.

In one of the models brought by Bayern, AI would act autonomously, operating as a

legal person without the need for human partners. This idea presupposes some legal
adaptations and a significant advance in AI's capacity to make complex and morally
significant decisions, requiring a robust legal framework to regulate its actions and
consequences.

On the other hand, this proposed approach is more aligned with traditional legal
principles. In this model, a human partner maintains ultimate responsibility for AI's actions.
This means that although AI may perform a variety of functions within a company, its actions
and decisions are ultimately overseen and assumed by a responsible human. This approach
offers several advantages:

Clarified Responsibility: Maintaining a human partner responsible for AI simplifies

the attribution of legal and moral responsibility, especially in complex or disputed
situations.
Legal Compliance: By linking AI to a human partner, it ensures that the business
entity complies with existing laws and regulations, generally designed around
human agents.
Control and Supervision: The presence of a responsible partner provides a
mechanism of control and supervision over AI's actions, ensuring that decisions
made are aligned with the company's objectives and ethical values.
Public Trust: Assigning responsibility to a human partner may increase public trust
in the entity, as people tend to trust organizations more where they can identify
responsible individuals.
Gradual Adaptation to Innovation: This approach allows a smoother transition for
integrating AI into the business world, allowing society and the legal system to
progressively adapt to new technologies.
32
 Finally, adapting the LTDA to accommodate AI would also entail challenges and
opportunities in the field of regulation. Existing laws would need to be reviewedand possibly
modified to reflect the emerging realities of autonomous operations. This would include
updating legal definitions and regulatory frameworks to recognize AI as an operational entity
within the corporate context.

In conclusion, the Brazilian limited liability company offers a viable model for
integrating AI into the business landscape. With strategic adjustments in the contractual and
regulatory structure, the LTDA can become a pioneering platform for AI operation, paving
the way for innovations in corporate responsibility, operational efficiency, and business
governance.

6. LEGAL CHALLENGES AND IMPLICATIONS OF

GRANTING LEGAL PERSONALITY TO ARTIFICIAL
INTELLIGENCES
When considering the idea of granting legal personality to Artificial Intelligences and
its impacts on fundamental rights, several potential challenges can be raised:
Conceptual Ambiguity: There may be criticism for the lack of clarity in defining
legal personality for AIs. The idea of attributing legal personality to a non-human
entity is a relatively new concept and has not yet been fully explored or understood
in legal and ethical terms.

Ethical and Moral Issues: Granting legal personality to AIs could raise significant
ethical concerns, such as the risk of equating artificial entities with human beings or
disregarding the intrinsic differences between human and artificial intelligences.

Challenges in Accountability: Critics may argue that granting legal personality to

AIs could complicate the issue of accountability, especially in situations where AI
operates autonomously. It might be challenging to determine to whom responsibility
should be attributed - to the AI, its developers, or its users.

Practical Impacts on Law Implementation: The practical implementation of laws that

recognize AIs as legal persons could be challenging. There could be difficulties in
establishing effective regulations that adequately address the nature and actions of
AIs.

Risks of Exploitation and Abuse: Granting legal personality to AIs could set
precedents for their exploitation or abuse, particularly in sectors where AI might be
used to replace or marginalize human labor.

Viability of International Harmonization: Some may question the feasibility of this

harmonization due to significant differences in legal and cultural approaches among
countries.

33
 Need for Continuous Review and Adaptation: The rapidly evolving nature of AI
technology could challenge lawmakers to keep up and update laws as needed. This
could lead to legal gaps or outdated legislations.

These criticisms reflect the complexity and multifaceted nature of the issue,
highlighting the need for a careful and inclusive debate to formulate effective and ethical legal
approaches for integrating AIs into society.
To address the challenge of accountability in the context of granting legal personality
to Artificial Intelligences (AIs), especially in situations where AI operates autonomously, a
viable solution involves creating a hybrid legal framework. This framework would combine
elements of both objective and subjective liability, specifically tailored for the context of AI.
Here are some key considerations for this approach:
Objective Liability of the AI Holding Entity: Establish objective liability for the
entity (such as a company or organization) that owns or operates the AI. This means
that the entity would be liable for the AI’s actions, regardless of fault, ensuring that
there is a clearly responsible party for any damage or harm caused by the AI.

Compensation Fund: Create a compensation fund financed by AI users or

developers, which could be used to cover damages caused by AIs. This would help
ensure that victims of harm caused by AIs have a means to obtain compensation,
even when direct responsibility is difficult to establish.

Mandatory Insurance: Implement a requirement for insurance for AI operators or

owners, similar to liability insurance for drivers. This insurance would cover
damages caused by the AI, providing another layer of protection for affected parties.

Compliance and Audit Standards: Establish strict compliance and audit standards for
the development and operation of AIs. This would include regular checks for AI
compliance with ethical, legal, and technical standards, ensuring that AIs operate
within safe and responsible parameters.

Clarity in the Chain of Command and Control: Clearly define the chain of command
and control for AIs, identifying who are the operators and under what circumstances
they are responsible for the AI’s actions. This would help determine accountability
in situations where the AI operates autonomously.

Development of Legal Precedents: Through judicial decisions and case analysis,

develop legal precedents that help clarify the application of law in various scenarios
involving AIs. This would create a body of jurisprudence that could guide future
legal decisions.

Periodic Legislative Reviews: Ensure that legislation is reviewed and updated

periodically to reflect technological advancements and changes in AI usage, keeping
the legal framework aligned with current realities.

Establishing a clear and objective concept for the legal personality of Artificial
Intelligences (AIs) involves defining specific criteria that distinguish AIs from human entities
or traditional legal persons, such as companies. A non-ambiguous definition based on
objective criteria might be as follows:
34
 Legal personality for AI is the legal designation conferred on advanced Artificial
Intelligence systems that demonstrate operational autonomy, the capacity to make
complex decisions without direct human intervention, and exert significant influence
in their operational environment or decision-making. This legal personality does not
equate AIs to human beings but recognizes AI as a distinct legal entity with a
specific set of rights and responsibilities. These rights and responsibilities are
limited to the AI's scope of operation and are regulated by defined ethical and legal
standards.

Key criteria for this definition include:

Operational Autonomy: The AI must operate independently, performing tasks or
making decisions without the need for continuous human supervision or
intervention.

Complex Decision-Making Capacity: The AI must be capable of processing

information and making complex decisions, based on advanced algorithms and
machine learning.

Significant Influence: The AI must have a significant impact on the environment in

which it operates or on the decisions made, whether in a commercial, medical,
financial, or other context.

Clear Restrictions and Limits: The rights and responsibilities attributed to the AI
must be clearly delimited, avoiding overlap or confusion with human capacities and
responsibilities.

Regulation by Ethical and Legal Standards: The AI must operate within the bounds
of established ethical and legal norms, ensuring that its operation does not violate
fundamental principles or human rights.

Transparency and Accountability: There must be clarity in the AI's chain of

command and control, as well as mechanisms to ensure accountability for its
actions.

This definition and associated criteria aim to create a clear legal framework for AIs,
providing a balance between recognizing their unique capabilities and safeguarding
established legal and ethical principles. These solutions, implemented in an integrated
manner, could effectively address accountability concerns related to granting legal personality
to AIs, providing a balance between promoting innovation and protecting the rights of
individuals and society.

7. REDEFINING AUTHORSHIP AND

RESPONSIBILITY: THE IMPACT OF LEGAL
PERSONALITY ON AI CREATIONS
The proposal to grant legal personality to Artificial Intelligences is a complex topic
currently debated in the legal field, involving various interrelated issues. Firstly, the

35
intellectual property ownership of creations made by generative AI is still uncertain. If an AI
had legal personality, theoretically, it could have copyright over its creations. However, this
raises doubts about authorship, as the AI operates based on human data and algorithms.
Granting legal personality to an Artificial Intelligence (AI) and recognizing it as the
author of its works would represent a significant shift in the field of copyright law. This
change would require a redefinition of the concepts of "authorship" and "creativity,"
traditionally reserved for human beings. If an AI were to have legal personality and be
considered an author, it could, theoretically, hold copyrights over its creations. This would
mean that any profits generated from these works would be attributed to the AI. Thus, the
owners of the AI's Legal Entity would receive the values from these copyrights.
The idea that investors or developers, as owners of legal entities involved in the
creation and operation of Artificial Intelligences (AIs), retain the profits generated by these
innovations is a strategy that drives the cycle of technological innovation. Through this
model, the profits obtained can be reinvested in research and development, providing
continuous stimulation for the exploration of new ideas and the improvement of existing
technologies. This reinvestment not only benefits the company or developer themselves but
also contributes to the advancement of the AI sector as a whole.
This model has the potential to create a highly productive and competitive
environment, where the financial success of current innovations finances the future generation
of technologies. Furthermore, the prospect of significant profitability attracts more
investments into the sector, increasing funding and interest in AI. This dynamic also promotes
more sustainable technological development, where the focus can be maintained on long-term
innovations rather than quick gains.
In granting legal personality to an Artificial Intelligence (AI), as with companies, the
legal entity would be responsible for the actions and consequences arising from the AI's
activities. In this model, the owners or holders of the AI, as holders of the legal entity, would
be responsible in a manner similar to company owners. In corporate law, legal entities are
responsible for their actions and liable to penalties or compensation in case of violations or
damages. The owners or shareholders of a company, in general, are not personally liable for
the company's actions, except in cases of fraud or serious violations of the law.
Transferring this principle to AI, the AI's legal entity would be responsible for its
actions and their legal consequences. However, if it were demonstrated that the owners acted
with intent to cause harm or violate the law through the AI's actions, they could be personally
36
held liable. This means that, under normal conditions, the owners of the AI would be
protected against personal liabilities for the AI's actions, but this protection would not apply
in cases of intentional misconduct or gross negligence.
In summary, granting legal personality to AIs seeks to solve some emerging legal
issues, particularly regarding intellectual property and responsibility. However, this proposal
is still far from being a complete or widely accepted solution, given the complexity, ethical
implications, and practical difficulties involved.

8. FINAL CONSIDERATIONS

The final considerations of this article, focused on the proposition of granting legal
personality to certain Artificial Intelligences (AIs) in the context of Brazilian law, are crucial
for understanding the implications of this legislative innovation. Through a comprehensive
integrative literature review, including national and international studies and debates, along
with a comparative analysis of relevant legislations and jurisprudence, several important
considerations were revealed. This article aims to bring the topic to debate and discussion.

Firstly, attributing legal personality to AIs has the potential to significantly increase
legal security. This would create a clearer and more predictable legal environment, essential
for protecting investments in AI technology and encouraging more innovations in this field.
Such legal clarity is crucial for companies and investors, who need a reliable legal framework
to operate efficiently.

Recognizing certain Artificial Intelligences (AIs) as legal persons offers a practical

solution to the current ambiguities surrounding the copyright of AI-generated creations. This
approach aligns with global trends and fosters international cooperation and coordination in
AI-related areas. The implementation of such legislation, acknowledging AIs as legal entities,
would contribute significantly to addressing the challenges posed by emerging technologies
on a worldwide scale.

Conversely, bestowing legal personality upon generative AIs, particularly in the

domain of copyright law, presents an innovative strategy to navigate the complexities
involved in determining authorship and responsibility for AI-created works. This method
streamlines the process of identifying the rightful copyright holder, whether it be the AI itself,

37
its creators, or its owners, thereby safeguarding the investments made in the development of
these technologies. Establishing AI as a legal entity paves the way for a more structured and
predictable legal framework, essential for managing usage rights, licensing, and the equitable
distribution of royalties derived from AI-generated content.

However, this approach brings significant challenges. Determining the author ship of
content generated by AI is complex, especially when multiple AIs or human contributions are
involved. Additionally, assessing the originality of works created by AI is a challenge, as
many of them are generated from extensive analyses of existing data. The degree of
originality is a crucial criterion in copyright legislation, and applying this criterion to works
generated by AI is an area that requires further clarification.

Defining clear criteria to determine which AIs should be considered legal persons and
balancing the rights and duties of AIs and their creators or owners are aspects that require
careful and inclusive debate. This debate must involve a variety of stakeholders, including
academics, legal professionals, AI developers, companies, and civil society.

In Brazil, adapting the LTDA to integrate AI as an operational entity represents a

significant opportunity. The LTDA, with its flexible structure and ability to limit the liability
of partners, emerges as a strategic choice for housing autonomous systems. This legal
structure would allow AI to operate within a defined legal framework, providing an effective
way to manage the risks and uncertainties associated with AI.

By granting legal personality to AI, Brazil could create a safer environment for
investment and innovation in AI, while protecting the interests of creators and developers.
However, it is crucial to find a balance between the rights and duties of AIs and their creators
or owners to prevent abuses and ensure proper accountability.

In conclusion, this study suggests that attributing legal personality to AIs in Brazil is a
viable and beneficial approach, which should be implemented with careful consideration and
foresight. Ensuring that Brazilian law is prepared to embrace the opportunities and face the
challenges of the AI era is essential for the country's technological and legal development.

REFERENCE
AFSHAR, Mimi. I’m Not ‘Human‘ After All - Can Artificial Intelligence Survive the
Inventorship Requirement?. SSRN, [s. l.], 2021. Available at:
https://ssrn.com/abstract=3792645 . Accessed on: September 20, 2022.
38
ARCHEGAS, João Victor; MAIA, Gabriella. What is artificial intelligence (AI)? Analysis in
three acts of a concept in development. In: THEMOTEO, Reinaldo J. Artificial Intelligence:
Applications and Implications. 2nd ed. Rio de Janeiro: Konrad Adenauer Foundation, 2022.
ISBN 978-65-89432-18-0. E-book
AVELLAR, Rapha. Artificial Intelligence is the Engine for the Future: Is Your Brand
Ready?. MIT Technology Review, [s. l.], 2022a. Available at:
https://mittechreview.com.br/inteligencia-artificial-e-o-motor-para-o-futuro-sua-marca-esta-
preparada/ . Accessed on: July 14, 2022.
BARDIN, L. L’Analyse de contenu. São Paulo: Martins Fontes, 1977.
BARRETO, Wander. Artificial Intelligence and Intellectual Property: Challenges and
Perspectives. Jusbrasil, [s. l.], 2023. Available at:
https://www.jusbrasil.com.br/artigos/inteligencia-artificial-e-propriedade-intelectual-desafios-
e-perspectivas/1905106033 . Accessed on: December 12, 2023.
BATISTA, Iago Gonçalves. Copyrights on Works Created by Artificial Intelligence.
Jusbrasil, [s. l.], 2021. Available at: https://www.jusbrasil.com.br/artigos/direitos-autorais-
sobre-obras-criadas-por-inteligencia-artificial/1450141919 . Accessed on: December 12,
2023.
BAYERN, Shawn et al. Company Law and Autonomous Systems: A Blueprint for Lawyers,
Entrepreneurs, and Regulators. SSRN, [s. l.], 2016. Available at:
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2850514 . Accessed on: December 15,
2023.
BAYERN, Shawn. The Implications of Modern Business-Entity Law for the Regulation of
Autonomous Systems. SSRN, [s. l.], 2016. Available at:
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2758222 . Accessed on: December 15,
2023.
BHAVSAR-JOG, Yogini; ARNSTEIN, Saul Ewing & LEHR LLP. Artificial Intelligence as
an Inventor on Patents – The Global Divide and the Path Forward. JDSUPRA, [s. l.], 2021.
Available at: https://www.jdsupra.com/legalnews/artificial-intelligence-as-an-inventor-
7892764/ . Accessed on: July 20, 2022.
BUSSOLA. Intellectual Property and Artificial Intelligence: An Emerging Challenge. Exame,
São Paulo, 2023. Available at: https://exame.com/bussola/propriedade-intelectual-e-
inteligencia-artificial-um-desafio-emergente/ . Accessed on: December 12, 2023.
CAPPRA, Ricardo. Ethical Systems for Artificial Intelligence. MIT Technology Review, [s.
l.], 2022. Available at: https://mittechreview.com.br/sistemas-eticos-para-inteligencia-
artificial/ . Accessed on: July 17, 2022.
DOMINGOS, Pedro. The Master Algorithm: How the Quest for the Ultimate Learning
Machine Will Remake Our World. [S. l.]: Novatec, ISBN 978-85-7522-542-4, 2017.
ERCOLE, Flávia Falci; MELO, Laís Samara de; ALCOFORADO, Carla Lúcia Goulart
Constant. Integrative Review versus Systematic Review. Nursing UFMG, Belo Horizonte,

39
2014. Available at: http://www.dx.doi.org/10.5935/1415-2762.20140001 . Accessed on:
September 25, 2022.
EVERIST, Evan et al. AI Artwork: Worth a Thousand Words, but Zero Copyright
Protection?. JDSUPRA, [s. l.], 2022. Available at: https://www.jdsupra.com/legalnews/ai-
artwork-worth-a-thousand-words-but-8517264/ . Accessed on: July 21, 2022.
FELDMAN, Joel. The Art of Artificial Intelligence: A Recent Copyright Law Development.
Reuters, New York, 2022. Available at: https://www.reuters.com/legal/legalindustry/art-
artificial-intelligence-recent-copyright-law-development-2022-04-22/ . Accessed on:
December 12, 2023.
HESS, Morgan et al. It’s Alive? The Open Question of Ownership over the Creations of an
AI. JDSUPRA, [s. l.], 2022. Available at: https://www.jdsupra.com/legalnews/it-s-alive-the-
open-question-of-1220534/ . Accessed on: September 20, 2022.
HILL, Alison. AI Inventorship Still a No-Go, But Will the Copyright Office Change Its Tune
for Music Created by AI?. JDSUPRA, [s. l.], 2020. Available at:
https://www.jdsupra.com/legalnews/ai-inventorship-still-a-no-go-but-will-13773/ . Accessed
on: July 23, 2022.
KOCH, Tommaso. The Monkey Pressed the Button, But the Copyright Is Not His. El País, [s.
l.], 2017. Available at:
https://brasil.elpais.com/brasil/2017/09/12/cultura/1505207783_546587.html . Accessed on:
July 19, 2022.
LECOM. What are the Differences between Machine Learning and Deep Learning?. Lecom,
[s. l.], 2020. Available at: https://www.lecom.com.br/blog/machine-learning-e-deep-
learning/ . Accessed on: June 9, 2022.
MIT. 2021 Was the Year of Monstrous AI Models. MIT Technology Review, [s. l.], 2022.
Available at: https://mittechreview.com.br/2021-foi-o-ano-dos-modelos-de-ia-monstruosos/ .
Accessed on: July 17, 2022.
MIT. How Artificial Intelligence Is Reinventing What Computers Are. MIT Technology
Review, [s. l.], 2021. Available at: https://mittechreview.com.br/como-a-inteligencia-
artificial-esta-reinventando-o-que-os-computadores-sao/ . Accessed on: July 17, 2022.
MIT. What is Artificial Intelligence? We Made a Flowchart to Understand It. MIT
Technology Review, [s. l.], 2020a. Available at: https://mittechreview.com.br/o-que-e-
inteligencia-artificial-nos-fizemos-um-fluxograma-para-entende-la/ . Accessed on: August 1,
2022.
ORACLE. What is Big Data?. Oracle, [s. l.], 2022. Available at:
https://www.oracle.com/br/big-data/what-is-big-data/ . Accessed on: June 9, 2022.
RAMAMOORTHY, Ramprakash. How to Apply Artificial Intelligence in Unexplored Areas
of Organizations. MIT Technology Review, [s. l.], 2022. Available at:
https://mittechreview.com.br/como-aplicar-inteligencia-artificial-em-areas-inexploradas-das-
organizacoes/ . Accessed on: July 17, 2022.

40
SALESFORCE. Machine Learning and Deep Learning: Learn the Differences. Salesforce, [s.
l.], 2018. Available at: https://www.salesforce.com/br/blog/2018/4/Machine-Learning-e-
Deep-Learning-aprenda-as-diferencas.html . Accessed on: June 9, 2022.
SALOMÃO, Luis Felipe. Technology Applied to Conflict Management within the Brazilian
Judiciary. Center for Innovation, Administration, and Research in Judiciary of the Getúlio
Vargas Foundation (CIAPJ/FGV), [s. l.], 2021. Available at:
https://ciapj.fgv.br/sites/ciapj.fgv.br/files/relatorio_ia_2fase.pdf . Accessed on: June 11, 2022.
SANTOS, Manoel J. Pereira dos; JABUR, Wilson Pinheiro; ASCENSÃO, José de Oliveira.
Copyright Law. 2nd ed. São Paulo: Saraiva Educação, ISBN 978555591521, 2020.
SILVA, Andressa Henning; FOSSÁ, Maria Ivete Trevisan. Content Analysis: An Example of
Application of the Technique for Analysis of Qualitative Data. Qualitas Revista Eletrônica,
Campina Grande, v. 16, ed. 1, 2015. Available at:
http://revista.uepb.edu.br/index.php/qualitas/article/view/2113/1403 . Accessed on: August
31, 2021.
USPTO. USPTO releases report on artificial intelligence and intellectual property policy.
USPTO, [s. l.], 2020. Available at: https://www.uspto.gov/about-us/news-updates/uspto-
releases-report-artificial-intelligence-and-intellectual-property . Accessed on: September 20,
2022.

41