Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

COPYRIGHT LAW AND ARTIFICIAL INTELLIGENCE

GENERATED WORKS: OWNERSHIP AND LIABILITY

Abhimanyu Bhargava, UPES, Dehradun

ABSTRACT

The emergence of Artificial Intelligence (AI) has brought new challenges to

copyright law, particularly with regard to ownership and liability of AI-
generated works. As AI technology advances, machines are becoming more
capable of generating creative works such as music, literature, and art,
raising important questions about the rights of ownership for these works.
This research paper explores the current state of copyright law with respect
to AI-generated works, including the key legal issues and challenges that
arise in determining ownership and liability. The study analyzes the various
approaches taken by different jurisdictions around the world in dealing with
the ownership and liability of AI-generated works, and examines the role of
existing copyright laws in addressing these challenges. Additionally, the
paper discusses the potential implications of the emerging legal landscape on
the creative industries, AI developers, and other stakeholders. The research
concludes with recommendations for policymakers, legal practitioners, and
other relevant actors to address the legal challenges posed by AI-generated
works, and to ensure the fair and equitable distribution of ownership rights
and liabilities.

Page: 1
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

Table of Contents:

1. A description of copyright legislation and how AI-generated works are affected by it

2. A description of and traits for AI-generated works

3. The contribution of human creativity to works produced by AI

4. The question of who holds the copyright for AI-generated works.

5. The part AI programmers and developers play in property ownership

6. AI's responsibility for copyright violations

7. Humans' responsibility for AI-generated works

8. Global perspectives on copyright legislation and artificial intelligence-generated works

9. Difficulties and consequences for copyright legislation in the era of AI-generated works

10. Potential answers to the ownership and liability problems associated with AI-generated
works.

A description of copyright legislation and how AI-generated works are affected by it

A legal framework known as copyright law gives writers the sole authority to manage and
safeguard their original works of authorship. By assuring that creators are compensated for
their efforts and have control over how their works are used, distributed, and monetized,
copyright law seeks to promote creativity and innovation.

The use of copyright legislation to protect AI-generated works has drawn more attention as
Artificial Intelligence (AI) technology has advanced. Artificial intelligence (AI) generates art
devoid of any direct human involvement. Music, visual arts, written works, and other types of
artistic works may be included.

AI-generated works pose several legal issues in the context of copyright law, including who
owns the copyright in the work and whether the work qualifies for copyright protection.
According to copyright law, a product must be original and permanently fixed in a tangible

Page: 2
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

form of expression in order to be protected. This requirement raises issues regarding the
originality of a work produced completely by an algorithm and the validity of the algorithm as
a tangible form of expression1.

Determining the degree of human participation in the creative process is one of the key
challenges in applying copyright law to AI-generated works. In general, copyright law
acknowledges human authors as the initial creators of works. However, there may have been
little to no direct human participation in the creation of works produced by AI. In these
situations, it is unclear who should be credited as the work's creator.

Determining the level of copyright protection that should be offered to works created by AI is
another problem. The duration of copyright protection is usually set, after which the work
becomes public domain and may be used by anyone without permission. However, the
longevity of AI-generated works may differ from those produced by human writers, raising
concerns about how long copyright protection for these works should last.

Legal scholars and policymakers have suggested a number of solutions to handle these
problems, such as extending the definition of "authorship" to include algorithms or recognising
AI as a "coauthor" of the work. Another suggestion is to create a new class of intellectual
property rights known as "machine-generated copyrights" for AI-generated works.

In conclusion, copyright law is a complicated body of law with significant ramifications for
writings created by AI. It is possible that copyright law will need to be modified as AI
technology develops to handle the particular issues these works present.

A description of and traits for AI-generated works

Artificial intelligence-generated works are those that are produced by computer programmes
or artificial intelligence algorithms without any involvement from humans. Among these works
are poetry, novels, works of art, and other forms of artistic expression.

The fact that AI-generated works are completely produced by algorithms or computer
programmes is one of their key characteristics. There is no direct human participation in the

1
Colin R. Davies, “An Evolutionary Step in Intellectual Property Rights - Artificial Intelligence and Intellectual
Property” (2011) 27(6) Computer Law & Security Review 601-619, p. 617.

Page: 3
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

creative process, in contrast to works written by human writers. Instead, the algorithms create
works based on patterns and data analysis using machine learning and other AI methods.

Another feature of works produced by AI is that they can be produced much more quickly than
those produced by people. In a matter of seconds or minutes, AI algorithms can analyse
enormous quantities of data and produce new works based on that data. This can be especially
helpful in disciplines like music and the arts where quickness and effectiveness are valued.

The products produced by AI can also be very diverse and varied. Depending on the
information and specifications given to the algorithms, they are capable of producing works in
a variety of styles and genres. As a result, there may be a wider range of artistic expression and
new opportunities for creators and artists2.

However, a drawback of AI-generated art is that it sometimes lacks the subjective and
emotional elements that are frequently connected to human ingenuity. Technically sound works
produced by AI algorithms are possible, but they might lack the emotional depth or sense of
connectedness that comes from human authorship.

The ownership and copyright of works produced by AI are also a topic of debate, as is the
possibility that AI systems could be held accountable for copyright violation. These questions
are still being discussed and investigated as technology advances and the use of AI-generated
works increases.

The contribution of human creativity to works produced by AI

In the context of AI and the creative industries, there is ongoing conversation and debate about
the place of human creativity in works produced by AI. There is some debate regarding how
much human creativity is engaged in the production of art, music, literature, and other forms
of creative expression when AI algorithms are used.

One theory holds that works produced by AI are simply the outcome of a code or programme
following a set of guidelines. According to this viewpoint, human ingenuity played little to no
part in the production of these works. In other words, AI-generated artworks are not genuinely

2
Ysolde Gendreau, “Copyright Ownership of Photographs in Anglo-American Law” (1993) 15(6) European
Intellectual Property Review 207-211, pp. 210-211.

Page: 4
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

"creative" in the sense that they lack the same arbitrary, sentimental, or aesthetic decisions that
are typically associated with human creative expression.

Another viewpoint, however, holds that human ingenuity is crucial to the production of works
produced by AI. This viewpoint recognises that while the works themselves may be produced
by AI algorithms, the rules and guidelines that govern the creative process were largely
established by the people who developed and trained the algorithms3. Even though it is not
actively involved in the creation of each unique work, human creativity is nevertheless present
in the design and development of the AI system4.

Some 4people also contend that human creativity is evident in the choice, curation, and
interpretation of works produced by AI. Humans are increasingly needed in the creative
industries as AI-generated works proliferate to analyse and evaluate them, choose the most
intriguing and valuable ones, and give them context and meaning5. This calls for a different
kind of creativity, one that is concentrated on comprehending and valuing the distinctive
characteristics and potential of works produced by AI.

In conclusion, human creativity plays a complex and multifaceted part in works produced by
AI. Although these works may have been produced by AI algorithms, human creativity was
still used in the selection and interpretation of the works as well as in the design and growth of
the algorithms. The connection between human creativity and AI-generated works will
probably continue to change as long as AI technology does.

The question of who holds the copyright for AI-generated works.

The question of who should be regarded as the creator or owner of AI-generated works is
complicated and contentious, and it is not always obvious who should be so designated. The
author or creator of a work is typically given ownership of it under copyright law, but in the
case of works produced by AI, there is no obvious author or creator in the conventional sense.

In some circumstances, the owner of the AI system or algorithm that produced the piece may
be regarded as the copyright owner. This would imply that the owner of the copyright is the

3
Megan Sword, “To Err Is Both Human and Non-Human” (2019) 88(1) UMKC Law Review 211- 233, p. 213.
4
Ibid.
5
E.g. Joshua L. Simmons, “Inventions Made for Hire” (2012) 2(1) New York University Journal of Intellectual
Property and Entertainment Law 1-50, pp. 43-47.

Page: 5
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

business or person who built or owns the AI system that produced the work, regardless of
whether they were directly involved in its creation.

As an alternative, some contend that the owner of the copyright should be the person who first
gave the data or inputs to the AI system6. This would imply that, even if they had no direct
involvement in the creation of the work itself, the copyright ownership is based on the
individual who provided the original creative spark or inspiration that resulted in the creation
of the work.

Additionally, there is an increasing case for considering AI-produced works as a joint effort
between humans and machines, with each party making some kind of contribution to the end
result. This would imply that the AI system and the people who worked on its creation and
development should each own a portion of the copyright.

Through specialised legislation or regulations, some legal systems have tried to resolve the
problem of ownership of AI-generated works. For instance, the recently passed Copyright
Directive in the European Union contains clauses regarding ownership of AI-generated works,
which state that the person or group that "made the arrangements necessary for the creation of
the work" should be regarded as the owner of the copyright”7.

In conclusion, the question of who owns works produced by artificial intelligence is still up for
dispute. It is likely that new legal and ethical frameworks will arise to address these issues and
decide the proper ownership and copyright protections for AI-generated works as AI
technology continues to develop and become more pervasive in the creative industries.

The part AI programmers and developers play in property ownership

It is a complicated matter how developers and coders of AI relate to copyright ownership of

AIgenerated works. Although they are not usually regarded as the authors of AI-generated
works, AI developers and programmers are crucial to the creation of the AI systems and
algorithms that produce these works.

6
Richard Arnold, “Content Copyrights and Signal Copyrights: The Case for a Rational Scheme of Protection”
(2011) 1(3) Queen Mary Journal of Intellectual Property 272-279, p.
7
Jyh-An Lee, “Copyright Divisibility and the Anticommons” (2016) 32(1) American University International
Law Review 117-164, pp. 124-130.

Page: 6
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

If they played a major part in designing and developing the AI system or algorithm that
produced the work, the AI developers and programmers may occasionally be regarded as the
owners of the copyright in the AI-generated works. For instance, if the AI system was created
with a specific creative goal in mind, like composing music or creating art, and the developers
and programmers were heavily involved in the system's design and development, they may be
regarded as the rightful owners of the copyright in the works the system produces.

However, in other circumstances, the AI developers and programmers might not have played
a major part in the creation of the works produced by the AI system and might not be regarded
as the rightful owners of those works' copyright. For instance, the developers and programmers
may not be regarded as the owners of the copyright in any resulting creative works if the AI
system was created to analyse and process data to generate insights or predictions and the
resulting output was not meant to be creative works.

It is significant to note that contracts between the AI developers, programmers, and any third
parties engaged in the creation or use of the works may also have an impact on who owns the
copyright to AI-generated works. The terms of the collaboration agreement, for instance, may
decide who owns the copyright to the works produced by an AI system created through a
collaboration between a business and an outside AI development team.

As a result, the involvement of AI programmers and developers in determining who owns the
copyright to works produced by AI is nuanced and situation-specific8. While they might be
regarded as the copyright owners in some circumstances, their ownership rights might be
restricted or shared based on their particular contribution to the design and development of the
AI system as well as any contracts or other applicable legal requirements.

AI's responsibility for copyright violations

The legal system has not yet completely addressed the complicated problem of AI's
responsibility for copyright infringement. Since AI is basically a tool or technology used to
create works, it cannot be held liable for copyright infringement because it is not regarded as a
legal person. However, there is still a possibility that the actions of AI could result in copyright

8
Richard S. Murphy, “Property Rights in Personal Information: An Economic Defense of Privacy” (1996) 84(7)
Georgetown Law Journal 2381-2418, p. 2412.

Page: 7
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

infringement, and there are various legal theories and methods that can be applied to handle
this problem.

Holding the human users or operators of the AI system accountable for any possible violations
is one way to deal with the issue of AI's liability for copyright infringement. This strategy is
predicated on the notion that, even though the infringing work may have been produced by an
AI system, it is ultimately the user or operator who is in charge of deciding how to use the
system and making sure that it conforms with copyright laws.

Another option is to hold the creators or producers of the AI system accountable for any
property violations that may arise from the use of the system. This strategy is predicated on the
notion that the designers or producers of the AI system have a duty to make sure that their
technology is developed and used in a manner that respects the rights of others9. Additionally,
it is conceivable that modifications to copyright legislation could be made to include liability
for infringement and particular provisions for AI-generated works. To clarify ownership and
liability concerns, some legal scholars have suggested, for instance, that a new legal category
be created for AI-generated works that is comparable to the idea of "work made for hire" in US
copyright law.

Overall, the problem of AI's liability for copyright infringement is complicated and constantly
changing, necessitating thorough analysis of the unique circumstances present. It is clear that
the expanding use of AI in the creative industries will continue to raise new legal and ethical
challenges, and that legislators will need to address them. While there are various legal
strategies that can be used to address this problem.

Humans' responsibility for AI-generated works

Another significant problem that has emerged with the increased use of AI in the production
of creative works is the liability of humans for works produced by AI. While AI-generated
works might not always be protected by copyright laws, they may still pose moral and legal
issues regarding responsibility and liability.

In general, there are many various ways that humans can be held accountable for works
produced by AI. The use of copyrighted content in the training data used to develop the AI

9
Megan Sword, “To Err Is Both Human and Non-Human” (2019) 88(1) UMKC Law Review 211- 233, p. 213.

Page: 8
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

system raises one potential liability problem. The human operators or developers of the system
could be held accountable for any ensuing copyright infringement if the AI system was trained
on copyrighted content without authorization10.

The application of AI to the creative process raises yet another possible liability concern. Even
though AI-generated works might not always be protected by copyright laws, they can still lead
to moral dilemmas regarding the place of human ingenuity in the creative process. For instance,
even if a human only offers very little guidance or input to an AI system that produces a piece
of work, they might still be regarded as the work's author and liable for any moral or legal
ramifications that follow.

In some circumstances, contracts or other legal factors may also have an impact on how
responsible people are for works created by AI. For instance, the terms of a business or
organisation's contract with an AI developer or operator may govern their responsibility for
any resulting legal or moral issues if they use an AI system to produce works.

In the end, the liability of humans for works produced by AI is a complicated and developing
problem that will need to be carefully taken into account by legal professionals and politicians.
It is crucial to make sure that legal frameworks are in place to address potential liability issues
and safeguard the rights of all parties concerned as AI continues to play an increasingly
significant role in the creative industries.

Global perspectives on copyright legislation and artificial intelligence-generated works

The issues of copyright law and AI-generated works have global implications and are viewed
from a variety of international viewpoints. There are several international agreements and
treaties that provide a foundation for defending the rights of authors and owners of copyrighted
works, despite the fact that copyright laws vary from one country to the next.

The Berne Convention for the Protection of Literary and Artistic Works is one international
accord that has bearing on copyright law and AI-generated works. The Berne Convention,
which contains provisions relating to ownership, duration, and enforcement of copyright, sets
minimum standards for copyright protection in member countries. The Berne Convention

10
J. Frankenfield, Artificial Intelligence (AI), March 13, 2020 [online] Available at:
https://www.investopedia.com/terms/a/artificial-intelligence-ai.asp

Page: 9
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

offers a framework for defending the rights of authors and owners of copyrighted works
regardless of how they were made, even though it does not explicitly address AI-generated
works.

The World Intellectual Property Organization (WIPO) Copyright Treaty is another

multinational agreement that is pertinent to copyright law and AI-generated works. The WIPO
Copyright Treaty contains provisions relating to the use of technological safeguards to protect
copyrighted works and offers additional protections for digital works11. The WIPO Copyright
Treaty offers a framework for safeguarding the rights of authors and owners of copyrighted
works in the digital age, even though it does not explicitly address AI-generated works.

A number of nations have also taken specific actions to handle copyright law and AI-generated
works in addition to these international agreements. For instance, the European Commission
has published guidelines on copyright in the digital common market in the European Union,
which include suggestions for handling problems involving AI-generated works. Similar to
this, the Copyright Office in the United States has published a report on how AI is affecting
copyright law. This report contains suggestions for how to handle ownership and liability
concerns with AI-generated works.

Overall, there are a number of international agreements and guidelines that provide a
framework for protecting the rights of creators and owners of copyrighted works, regardless of
how they were created, even though copyright law and AI-generated works are complex and
dynamic issues that differ from country to country.

Difficulties and consequences for copyright legislation in the era of AI-generated works

In terms of copyright law, the emergence of AI-generated works poses a number of difficulties
and consequences. Key issues and consequences include the following:

1. Authorship and ownership: The issue of who is the author and owner of AI-generated
works is a significant obstacle for copyright legislation. As was previously mentioned,
it is frequently unclear who the author or creator of an AI-generated work is, which can
lead to ethical and legal quandaries regarding property ownership.

11
The 7 Most Pressing Ethical Issues in Artificial Intelligence 2019 [online] Available at:
https://kambria.io/blog/the-7- most-pressig-ethical-issues-in-artificial-intelligence/

Page: 10
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

2. Copyright eligibility: Determining whether AI-generated works qualify for copyright

protection presents another problem for copyright law. The extent of copyright
protection in the age of AI is questioned by the possibility that some AI-generated
works are deemed to be too banal or unoriginal to be protected by the copyright.

3. Copyright infringement: Using AI to produce works can lead to problems with

copyright infringement. For instance, even if the AI-generated work is not an exact
replica of the original work, an AI system taught on copyrighted material may produce
works that violate the author's copyright.

4. Liability: In the period of AI-generated works, liability is yet another significant issue
for copyright law. As was mentioned earlier, it is frequently uncertain who is
responsible for any legal or moral issues that result from AI-generated works, which
can pose problems for copyright law.

5. Fair use and transformative works: Concerns about fair use and the extent of copyright
protection are raised when AI is used to create transformative works or works that
contain protected content. It is unclear whether using an AI system to produce a work
that transformatively incorporates copyrighted content would be regarded as fair use
under copyright law.

The emergence of AI-generated works poses a variety of issues and has repercussions for
copyright law overall. Identifying title and authorship, copyrightability, infringement, liability,
and the extent of fair use and transformative works are some of these difficulties12. Copyright
law needs to adjust to these new issues and implications as AI plays an ever-more-significant
role in the creative sectors.

Potential answers to the ownership and liability problems associated with AI-generated
works.

The title and liability concerns pertaining to AI-generated works can be resolved in a number
of ways. Some of the most significant answers are:

12
Robert Yu, “The Machine Author: What Level of Copyright Protection Is Appropriate for Fully Independent
Computer-Generated Works?” (2017) 165(5) University of Pennsylvania Law Review 1245-1270, p. 1258.

Page: 11
Indian Journal of Law and Legal Research Volume V Issue II | ISSN: 2582-8878

1. Adopting new legislation or amending current legislation to address ownership and

liability concerns linked to AI-generated works is one potential solution. For instance,
some nations have suggested giving AI systems a new legal personhood category that
would grant them some level of legal accountability and ownership rights.

2. Contracts and licencing agreements: Contracts and licencing agreements can also be
used to determine who owns and is responsible for AI-generated works. This might
entail asking AI creators or users to sign contracts outlining ownership and liability
rights for works produced by AI, as well as any royalties or other payments that might
be owed to the works' original authors.

3. Reforming copyright laws to better address ownership and liability problems pertaining
to AI-generated works is a third option. Changes to the requirements for
copyrightability and the extent of copyright protection for works produced with the aid
of AI may be necessary to address this.

4. Attribution and metadata: Requiring AI-generated works to include attribution and

metadata that explicitly identify the original authors of the work, as well as any other
pertinent details about the AI system used to create the work, is another potential
remedy. By doing so, it might be possible to create ownership and liability rights for
works produced by AI and prevent copyright infringement.

5. Industry standards and best practises: The industry establishing standards and best
practises for addressing ownership and liability problems relating to AI-generated
works is a final potential solution. This could entail establishing industry-wide
protocols for licencing and crediting AI-generated works as well as building guidelines
for AI developers and users.

In general, there are a number of potential answers to the title and liability problems associated
with AI-generated works. The establishment of clear ownership and liability rights for AI
generated works may require a combination of legislative action, contracts and licencing
agreements, copyright law reforms, attribution and metadata, industry standards and best
practices, and contracts and licencing agreements. However, there is no one-size-fits-all
solution to this complex issue.

Page: 12