Open Access Research Article
AI AS A CREATOR: LEGAL CHALLENGES IN COPYRIGHT INFRINGEMENT AND FAIR USE UNDER INDIAN LAW.
Published Paper
Article Details
AI AS A CREATOR: LEGAL CHALLENGES
IN COPYRIGHT INFRINGEMENT AND FAIR USE UNDER INDIAN LAW.
AUTHORED BY - ELEENA ANIL VADAKKEKARA
(LLM scholar),
CO-AUTHOR - DR. ARADHANA SATISH NAIR
(faculty)
Christ (Deemed To Be University),
Bengaluru
Abstract
The enforcement of copyright law has become increasingly complex in
the digital era, raising significant challenges in protecting creators' rights.
This paper examines issues related to copyright infringement, fair use, and
legal liabilities across different jurisdictions, with a particular focus on
India. It explores the fair use doctrine in India, the United States, and the
European Union, analyzing key legal interpretations and distinctions that
influence copyright enforcement. The study also addresses secondary liability,
including vicarious and contributory infringement, in the context of
unauthorized digital content distribution. Additionally, recent legislative
changes, such as the EU’s Copyright Directive, are reviewed to assess their
impact on copyright regulation and intellectual property law. The paper further
evaluates the role of the World Intellectual Property Organization (WIPO) in
promoting international copyright standards. As digital content consumption
continues to rise, this study highlights the necessity of evolving legal
frameworks to balance the protection of intellectual property with public
access to information. The findings provide insights into potential policy
reforms that ensure greater clarity and adaptability in copyright law while
addressing emerging legal challenges.
Keywords: Copyright Law, Fair
Use, Infringement, Intellectual Property, Legal Framework, India,
Introduction
Artificial Intelligence (AI) has
become a transformative force across various domains, with its origins tracing
back to 1956 when John McCarthy[1]
first introduced the term. AI aims to replicate human intelligence in machines,
enabling them to learn, reason, and make decisions. One of the most significant
advancements in AI is machine learning, introduced by Arthur Samuel[2] in
1959, which allows machines to analyze vast datasets and make predictions
without explicit programming. Since the 1980s, AI-based systems such as IBM’s
Deep Blue, Kismet, Dragon Systems, and AlphaGo have emerged, demonstrating AI's
increasing sophistication. Today, AI applications extend across multiple
industries, including healthcare, finance, space exploration, and the creative
arts.
AI’s influence in the creative sector
is exemplified by projects like ‘The Next Rembrandt,’ which utilized AI to
analyze and replicate Rembrandt’s artistic style, and Google’s ‘Deep Dream
Generator,’ which produces unique artwork by blending existing images. These
developments highlight AI's expanding role in both functional and expressive
domains, where it engages in human-like creativity.
Arend Hintze categorizes AI systems
into four stages: reactive machines, limited memory, theory of mind, and
self-awareness. Reactive machines, such as IBM’s Deep Blue, operate without
memory and respond solely to present stimuli[3].
Limited memory systems, like autonomous vehicles, rely on historical data for
improved decision-making. Although AI has yet to achieve ‘theory of mind’
capabilities, ongoing research seeks to develop systems that comprehend human
emotions and cognitive states[4].
Self-aware AI, representing the most sophisticated stage, remains an aspiration
for future technological advancements[5].
Despite AI’s vast potential, ethical
and legal concerns, particularly regarding copyright laws, have emerged. The
rapid digitization of content has made enforcing copyright protections more
complex. Copyright infringement occurs when copyrighted material is used
without authorization, leading to direct, indirect, or unintentional
violations. Copyright laws differ globally; some jurisdictions impose strict
penalties, while others consider factors such as intent and economic impact. A
key defense against infringement is the doctrine of fair use (or fair dealing
in some countries), which permits limited use of copyrighted material for
education, research, and journalism. In the United States, fair use is assessed
based on four factors: purpose, nature, extent, and market effect. Similarly,
India’s Copyright Act of 1957[6]
includes fair dealing provisions that allow certain exceptions.
A comparative analysis of copyright
enforcement highlights significant legal differences. The European Union’s 2019
legislation permits text and data mining for research, reflecting a balanced
approach to copyright protection and technological innovation. India’s legal
framework, influenced by British and American laws, attempts to accommodate
both the rights of content creators and public interests. Examining landmark
cases and legislative developments offers insights into how copyright laws are
adapting to technological advancements.
As AI evolves, concerns regarding its
training processes and potential copyright violations have grown. Machine
learning models require vast datasets, which may include copyrighted works,
raising legal compliance questions. AI training involves multiple stages, such
as data collection, preprocessing, and pattern recognition, where unauthorized
use of copyrighted content can result in infringement. Additionally, deepfake
technology exemplifies the ethical dilemmas posed by AI-generated content.
Deepfake models manipulate images, videos, or audio using adversarial training
techniques to produce realistic yet artificially generated content, often
without the consent of the original creators.The absence of comprehensive
regulations for AI-generated content necessitates a reassessment of copyright
laws to ensure a fair balance between intellectual property protection and
innovation. By exploring legal precedents, technological challenges, and
evolving regulations, this research aims to provide a comprehensive understanding
of copyright implications in the digital era.
Legal Research Questions:
(1) To what extent does Indian copyright law recognize AI-generated works as
eligible for copyright protection, and how does it compare with international
frameworks?
(2) What are the legal challenges and liabilities associated with copyright
infringement in AI-generated content under Indian law, and how do they align
with global trends?
(3)
How does the
fair use doctrine in India apply to AI-generated content, and what legal
reforms are necessary to address emerging copyright issues in AI training and
data mining?.
Research methodology
This study employs a doctrinal
legal research methodology, which involves an in-depth examination of statutes,
case laws, and international legal frameworks governing copyright
infringement and fair use under Indian law. A comparative legal
approach is adopted to analyze India’s copyright framework in relation
to the United States and the European Union, identifying key
similarities, differences, and areas for potential reform. The research is
based on primary sources, including the Copyright Act, 1957, decisions, and
relevant international treaties, supplemented by secondary
sources such as scholarly articles, reports from
the World Intellectual Property Organization (WIPO), and policy documents such
as India’s National Strategy for
Artificial Intelligence (NITI Aayog), provide contextual insights
into evolving regulatory trends. The study follows the Bluebook Citation, 20th Edition,
ensuring standardized referencing of legal sources. Based on the findings, policy
recommendations are proposed to enhance copyright protection while
fostering innovation within the AI-driven creative economy.
Copyright
Infringement and AI-Generated Works.
Machine learning processes vast
amounts of data, including potentially copyrighted content, leading to legal
concerns[7]. It
generally follows 3 kinds of Leaning: Supervised Learning which uses
structured, labeled data to produce expected outcomes. An Unsupervised Learning, where it
Analyzes unstructured data without predefined labels. And Finally, Reinforcement Learning which Adapts to
new environments by learning from past or existing decisions, data etc[8].
During this process of Training Of AI it or Date processing it undergoes
several stages which may includes Data
Collection: Gathering
extensive datasets; Data Cleaning: Eliminating errors, inconsistencies,
and biases;Data Formatting:
Standardizing data for compatibility; Pattern
Identification: Extracting
relevant insights and Training &
Evaluation: Dividing data into training and test sets for analysis.
2.1 AI Training and Copyright
Infringement Risks
The process of training artificial intelligence (AI) models raises
significant concerns about copyright infringement. Machine learning systems
often utilize unauthorized datasets, creating digital copies of existing
copyrighted works, which can be considered an act of reproduction[9].
During the training phase, AI models may engage in literal reproduction when
datasets are copied multiple times, leading to multiple instances of
potentially infringing copies[10].
Additionally, non-literal reproduction occurs when vast amounts of unauthorized
copyrighted content are used as input data, which can result in AI-generated
outputs that closely resemble the original material, potentially violating
copyright protections[11].
A notable example of potential copyright infringement in AI training
is deepfake technology[12]. Deep-fake
technology uses advanced neural networks to manipulate images, videos, and
music, creating content nearly identical to original works. It operates through
adversarial training, where a generator produces synthetic data while a
discriminator attempts to differentiate it from real content[13].
The process continues until the distinction becomes indistinguishable. Legal
concerns arise as deepfakes rely on copyrighted material, yet specific
regulations governing AI-generated works remain absent, making their legality
uncertain[14].
2.2 Legal framework for
infringement:
AI-generated content can potentially
infringe on third-party copyrights, trademarks, and even an individual's right
to privacy and publicity. Under U.S. law, copyright infringement by
AI-generated works is assessed based on two key factors: (a) whether the AI had
access to the copyrighted work, and (b) whether the output is substantially
similar to the original work.
A critical question that arises is
determining liability in cases of AI-generated infringement. Under current
legal doctrine, both the user and the AI developer or company may be held
liable. If a user is found directly responsible, the AI company may also be
held liable under the doctrine of vicarious infringement.
Liability
Indian copyright law:
Under Indian copyright law, secondary
liability for infringement can arise through either vicarious liability
or contributory infringement.[15] Vicarious
liability applies when an entity has the right and ability to control
the infringing activity and derives financial benefit from it[16]. Contributory
infringement occurs when an entity knowingly induces, causes, or
materially contributes to the infringement, even without directly engaging in
it[17].
While Indian courts have yet to
address liability for AI-generated works specifically, rulings on secondary
liability in digital contexts offer some guidance. In MySpace Inc. v. Super Cassettes Industries Ltd[18].,
the Delhi High Court ruled that MySpace was not liable for copyright
infringement by its users, as it complied with India's "safe harbor"
provisions, which protect intermediaries from liability if they implement
notice-and-takedown mechanisms and avoid involvement in content creation.
Conversely, in Christian Louboutin
SAS v. Nakul Bajaj & Ors[19].,
the court held that an e-commerce platform facilitating counterfeit sales was
liable for contributory infringement, as it had knowledge of the infringing
activity and materially contributed to it[20].
United States-An Established Doctrines of Liability:
In the United States, copyright
liability extends beyond direct infringement to include secondary liability,
which can arise under two key doctrines: Vicarious Liability which applies
when an entity has the right and ability to control infringing activity and
derives a financial benefit from it, even if they did not directly participate
in the infringement[21].And
Contributory Infringement that occurs when an entity knowingly
contributes to or facilitates infringing activity.
Beyond these traditional doctrines,
U.S. law also recognizes inducement liability, which
applies when a party actively promotes or encourages infringement[22].
This doctrine was established in the landmark case MGM Studios Inc. v. Grokster, Ltd. (2005)[23], where the U.S. Supreme Court held that
companies distributing peer-to-peer file-sharing software could be held liable
if they intentionally encouraged users to engage in copyright infringement. If
AI developers or platforms actively design their systems to enable or encourage
unauthorized copying, they may also be subject to this form of liability[24].
European Union:-Safe Harbor and Mere Conduit Protections:-
In the European Union, the E-Commerce Directive (2000/31/EC)
governs the liability of online intermediaries for copyright infringement. The
Directive establishes key exemptions for service providers: (i) Hosting Safe
Harbor – Platforms that host user-generated content (such as social media
and cloud storage services) are protected from liability, provided they do not
have actual knowledge of infringing content and act promptly to remove it when
notified[25]. (ii) Mere
Conduit Exception – Internet service providers (ISPs) and other
intermediaries that simply transmit infringing content without modifying it are
exempt from liability[26].
However, recent EU copyright reforms,
such as the Copyright Directive (2019/790[27]),
have introduced stricter obligations for digital platforms. Under Article 17,
platforms that facilitate content sharing may be required to implement content
recognition systems, making them more accountable for infringing material
uploaded by users. This raises an important questions about how the
AI-generated content will be regulated under these evolving legal standards?.
Fair use
USA:
Text and Data Mining (TDM) refers to
the computational process of extracting patterns and discovering new
information from large sets of unstructured data. This process often involves
analyzing text or other types of data through algorithms to find useful
insights. The United States lacks a specific statutory exception for Text and
Data Mining (TDM)[28].
However, TDM activities typically fall under the 'Fair Use Doctrine' in Section
107 of the U.S. Copyright Act[29].
Fair use is assessed using a four-factor test to determine whether a particular
use constitutes copyright infringement.
The four-factor test evaluates:
a) Purpose and character of the use –
it examines whether the use is commercial and whether it is transformative (i.e., whether it adds new
meaning, expression, or purpose to the original work). A transformative use weighs in favor of fair use,
while a commercial use weighs against it. However, if a work is highly
transformative, its commercial nature becomes less significant.
b) Nature of the Copyrighted Work:
Factual and non-fiction works are
generally granted less protection than creative works under copyright law. This
means that TDM involving factual data (e.g., research articles, databases,
public records) is more likely to be considered fair use than TDM involving
creative works like literature, art, or music.TDM often involves factual data,
which makes this factor favorable for those who argue that TDM activities
should be protected under fair use.
c) Amount and Substantiality of the
Portion Used:
This factor examines how much of the
copyrighted work is used in relation to the entire work. A limited,
non-substantial use is more likely to be deemed fair use.In the case of TDM,
the amount of copyrighted material used depends on the scope of the data
collected. Since TDM often involves the extraction of large amounts of data,
the use of substantial portions of copyrighted works may be necessary for the
analysis. However, the transformative nature of the use may outweigh the
quantity used.
d) Effect of the Use on the Market:
This factor evaluates the impact that
the use has on the potential market for the original work. If the use replaces
the original work in the market or causes economic harm to the copyright owner,
it is less likely to be considered fair use.TDM typically does not harm the
market for original works, particularly if the use is non-commercial and
transformative. However, the economic impact of using copyrighted works for
commercial purposes (e.g., training AI models) may need to be assessed on a
case-by-case basis.
Courts Through several case laws have provided significant guidance
on how digital reproductions of copyrighted
works which includes searchability or artistic
reinterpretation which constitutes fair use.
In Authors Guild, Inc. v. HathiTrust[30], the defendant, ‘HathiTrust’, created a full-text searchable database by scanning books
from university libraries. The digital collection allowed users to search for
specific terms without providing full access to entire books. The Second
Circuit ruled that digitizing books for
search functionality was transformative because the digitized
copies served a different function from the original works.
While the books were originally intended to be read in their entirety, the
digitized versions enabled users to search and analyze text data, making the
use distinct from the original purpose. This case reaffirmed that creating digital tools that enhance accessibility and research
capabilities can constitute transformative use under copyright law.
NXIVM Corp. v. Ross Institute[31]
& Oracle America, Inc. v.
Google LLC[32] cases established that an alleged
infringer’s bad faith or improper conduct does not automatically
preclude a fair use defense. Courts
ruled that fair use should be assessed based on the nature of the use itself, rather than the
intent behind it. This principle suggests that if a use meets the legal
criteria for fair use, it remains protected regardless of the defendant’s
motivations.
In Cartoon LP v. CSC Holdings[33]: The court ruled that non-volitional
uses of copyrighted works by computers, such as "intermediate operational
use," do not constitute infringement because they are not creative in
nature. In Burrow-Giles Lithographic
Co. v. Sarony[34]:
The court found that mechanical processes or transcoding of copyrighted works
are not considered expressive works and therefore do not infringe copyright.
Further in the case of Sega v. Accolade[35]:,
the court recognized non-expressive fair use, where copying functional code for
reverse engineering purposes was found to be fair use because the copying
involved functional elements, not creative ones. In Kelly v. Arriba[36]:
The court determined that using copyrighted images to create thumbnails for
search results was transformative fair use, as the purpose of the use was
different from the original purpose of the images.In another Case Of Authors Guild v. Google[37]:
Google’s scanning of books for keyword searches was deemed transformative fair
use, as the purpose was distinct from the original use of the books.
While the fair use doctrine protects
uses of copyrighted works in training technologies, it should be applied
judiciously to avoid diminishing the market value of the works. In Harper & Row Publ v. Nation Enters[38],
the court emphasized that fair use should not be interpreted too broadly,
especially in commercial contexts, to prevent harm to the copyright owner’s
rights.
When considering expressive uses of
works in training, such as creating music or advertisements, the legal
challenges become more complicated. Tools like Jukedeck, which use copyrighted
music in their training data, could harm the market for sound recordings by lowering
licensing rates and disrupting established business models[39].
There is currently a lack of clear legal guidance on regulating these uses,
making it crucial to strike a balance between protecting copyright owners'
rights and fostering innovation in content creation[40].
EUROPEAN UNION (EU):-
The European Union (EU) recognizes
the economic potential of Artificial Intelligence (AI) and aims to enhance its
competitiveness through increased research and development funding. However,
the EU's AI policy has largely overlooked intellectual property (IP) aspects[41],
which may explain its lag behind Asia and North America in private investment
and AI-related patents[42].
As AI development relies on large datasets, a flexible and industry-friendly
copyright regime is necessary to support data access for AI training. However,
tensions arise between promoting AI innovation and protecting creators of
original works used in AI datasets[43].
Text and data mining (TDM) is crucial
for machine learning, as it helps identify patterns from large datasets[44].
While works not under copyright can be freely mined, using copyrighted works
for TDM may infringe on the copyright holder’s reproduction rights[45].
There are differing views on whether TDM activities should lead to copyright
infringement liability[46].
Some argue that using copyrighted works for AI development shouldn’t be
considered fair use, particularly if the resulting AI output is commercially
exploited, as it could harm the copyright holder's market. Allowing AI
developers to use copyrighted works without liability could create inequitable
circumstances, potentially disadvantaging human creators. A balance must be
struck between the copyright holder's rights and the developer's needs[47].
The World Intellectual Property Organization (WIPO) currently offers no clear
guidance on whether using data from copyrighted works without permission
constitutes infringement, leading to different approaches in various
jurisdictions. Many countries have introduced exceptions in their copyright
laws for TDM, though these exceptions vary in scope[48].
Restrictions on the types of works, rights covered, commercial use, and
cross-border transfers, as well as requirements for lawful access, all limit
the application of these exceptions. Whether these restrictions are conducive
to AI growth remains to be examined.
Regarding their Legal Framework - The
European Union (EU) does not recognize the concept of fair use and
transformative use in the same way as the United States[49].
EU copyright law, particularly the 2001 Information Society Directive, sought
to harmonize copyright laws across Europe, especially concerning copyright
exceptions[50]. The
Directive allows for exemptions from copyright infringement for transient
copies created during technological processes, as long as the copies are
necessary and do not have independent creative value. These copies must also be
lawfully accessed. This exception applies to machine learning processes, where
temporary copies are made during training without being permanently stored.
However, the use of copyrighted works
under this exception must comply with a three-step test[51].
This test ensures the exception applies only in special cases that don't
conflict with the normal exploitation of the work or harm the legitimate
interests of the copyright holder. The first part of the test limits the
exception to specific purposes, while the second part requires that the use
does not interfere with the commercial exploitation of the work[52].
The third part ensures the use does not cause unreasonable loss to the
copyright holder, potentially requiring compensation[53].
In 2019, the EU introduced a specific
exception for text and data mining (TDM) under the Copyright in the Digital
Single Market Directive[54].
This new provision expanded the existing framework and allowed research
organizations and cultural heritage institutions to use the TDM exception for
scientific research. However, to apply this exception, researchers must
lawfully access the works being mined, typically through a contract, unless an
open access policy is in place.
Article 7[55]
of the Directive renders contractual clauses that limit the TDM exception
unenforceable. Article 4 [56]also
broadens the scope of the exception, allowing categories beyond research
organizations to use it for non-commercial purposes. One notable feature of the
2019 Directive is the reinforcement of the public domain: once a visual
artwork's copyright expires, its derivative materials are no longer protected
by copyright.
Despite these advancements, the 2019
Directive has some limitations, particularly for AI development[57]. Private
companies developing AI systems encounter barriers to TDM exception
implementation except through public-private partnerships that might impede
their capability to capitalize commercially from their developed products.
Natural rights holders under the Directive possess the option to deny mining
rights which potentially undermines innovation development. AI research teams
along with technology companies encounter implementation problems due to
undefined TDM parameters in the exception framework and diverse EU member state
interpretations. Several unresolved matters as well as ambiguity in the
Artificial Intelligence industry persist under the current Directive until
Europe achieves digital market unification.
Indian Perspective
The absence of a comprehensive legal
framework for regulating and supervising emerging technologies has led to
ambiguity in various legal aspects. In
Indian some of the initiatives, such as the policy paper released by NITI Aayog
titled National Strategy for Artificial Intelligence, . NITI Aayog
which has decided to focus on five
sectors that are envisioned to benefit the most from AI in solving societal
needs: a) Healthcare: increased
access and affordability of quality healthcare; b) Agriculture: enhanced farmers’ income, increased farm productivity
and reduction of wastage; c) Education:
improved access and quality of education;
d) Smart Cities and Infrastructure:
efficient and connectivity for the burgeoning urban population,ande) Smart
Mobility and Transportation: smarter and safer modes of transportation and
better traffic and congestion problems[58].
Additionally, a committee led by V. Kamkoti has been formed to promote research
and development in this domain and established the National AI Mission[59].The
policy emphasizes self-regulation, advocating for principles such as
transparency, privacy, equality, safety, inclusivity, and accountability.
However, uncertainty remains regarding the legality of Text and Data Mining
(TDM) and the use of data under existing copyright laws[60].
Section 52[61] of the
Copyright Act provides a detailed list of activities that qualify as fair
dealing and do not amount to copyright infringement, but it does not explicitly
address TDM. While Section 52(1)(a) permits the use of literary works for
private or personal purposes, including research, criticism, review, or news
reporting—subject to certain conditions—it does not extend to commercial use[62].
This principle was reinforced in Saregama India Ltd. v. Alkesh Gupta[63]and
Tips Industries Ltd. v. Wynk Music Ltd[64].,
where the courts clarified that fair dealing protections apply only to
non-commercial purposes, specifically for research and development, but not for
commercial exploitation[65].
To conclude, Copyright law is
constantly changing to keep up with digital advancements. The rise of
artificial intelligence, which relies on large amounts of data, has raised new
concerns about copyright infringement. Questions about who owns AI-generated
content, who is responsible for misuse, and how to balance innovation with
copyright protection remain unanswered. Different countries have taken
different approaches to these issues, leading to various legal interpretations.
In India, copyright law includes fair
use provisions under Section 52 of the Copyright Act, allowing limited use of
copyrighted material for education, research, and news. However, it does not
clearly address AI-generated content or text and data mining. In contrast, the
United States follows a flexible four-factor test to decide fair use, but court
rulings can vary. The European Union has taken a stricter approach with its
2019 Copyright Directive, requiring online platforms to take more
responsibility for copyright violations. Even with these regulations,
AI-generated content remains a gray area, and more legal clarity is needed.
The issue of liability is also
evolving. Digital platforms, AI developers, and internet service providers are
being held increasingly accountable for copyright violations. While some legal
protections exist, recent court rulings indicate that companies can be
responsible if they knowingly allow infringement. Courts have played a key role
in shaping these rules, and legal trends suggest stronger enforcement measures
in the future.
As technology advances, copyright
laws need to strike a balance between protecting creators' rights and promoting
public access to knowledge. The complexities of AI-generated works highlight
the need for updated legal frameworks. Policymakers and legal experts must work
together to create laws that support both innovation and copyright protection
in the digital world.
Conclusion
and Suggestions
Conclusion
The evolving landscape of copyright
law presents significant challenges in the digital era, particularly concerning
the intersection of legal frameworks and technological advancements. The study
highlights that while copyright laws have traditionally been designed to
protect human creators, they now face difficulties in addressing digital content
creation, unauthorized usage, and liability concerns. In India, fair use
provisions under the Copyright Act, 1957, provide certain exceptions for
education, research, and journalism. However, they require further clarity in
addressing digital reproduction, data mining, and unauthorized content
distribution.
A comparative analysis reveals that
different jurisdictions adopt varied approaches to copyright enforcement. The
United States applies a flexible fair use doctrine, allowing courts to
interpret cases based on key legal principles. Meanwhile, the European Union,
through the 2019 Copyright Directive, has imposed stricter obligations on
digital platforms to prevent copyright violations. Despite these advancements,
emerging digital challenges necessitate further legal refinements and
harmonization to ensure a balance between protecting intellectual property
rights and allowing legitimate public access to information.
The study further examines legal
liability, emphasizing the increasing responsibility of content-sharing
platforms, digital service providers, and intermediaries in cases of copyright
infringement. Judicial decisions continue to shape the enforcement landscape,
determining liability based on intent, economic harm, and technological impact.
As digital content consumption increases, the need for a comprehensive and
adaptive legal framework becomes imperative to safeguard intellectual property
without stifling access to knowledge and innovation.
Suggestions
5.2.i. Amending the fair dealing
provisions:
India’s fair dealing provisions under Section 52 of the Copyright Act
should be expanded to explicitly
define the scope of permissible uses of copyrighted works. The U.S.
fair use doctrine offers
courts discretion in assessing transformative
use, while Japan’s
approach allows limited content use for research and education.
India should introduce clearer statutory language distinguishing
permissible digital reproductions from unauthorized use, ensuring alignment
with evolving digital consumption patterns.
5.2.ii. Addressing Digital Content Liability (EU’s
Copyright Directive & U.S. Safe Harbor Law)
Digital platforms that facilitate
content sharing play a crucial role in copyright enforcement. The EU’s
Copyright Directive (Article 17) holds online platforms accountable
for unauthorized content distribution, while the U.S. Safe Harbor
provisions provide
liability exemptions for platforms that comply with takedown requests.
India should implement a tiered liability framework, ensuring digital
service providers take proactive measures to prevent copyright infringement
while preserving safe harbor protections for legitimate content-sharing
platforms.
5.2.iii. Digital Rights Management and Licensing
Mechanisms (South Korea’s Model)
South Korea has implemented a compulsory
licensing model that ensures copyright holders receive fair
compensation for the use of their works in digital spaces.
India should establish similar licensing mechanisms that promote
transparency in digital content distribution, ensuring that copyright owners receive
appropriate remuneration for content reproduction and public dissemination.
5.2.iv. 4. Enhancing Intermediary Regulations and
Compliance (EU Digital Services Act, US DMCA)
The EU Digital Services Act & U.S. Digital Millennium
Copyright Act (DMCA) regulate intermediary liability, ensuring
online platforms maintain accountability for copyright violations while
providing due process for content takedowns.
India should introduce clearer compliance obligations for online
intermediaries, requiring them to implement effective monitoring and takedown
mechanisms to prevent unauthorized content distribution.
5.2.v. Aligning
India’s Copyright Law with Global Standards (WIPO & International
Conventions)
The World Intellectual
Property Organization (WIPO) has
established international
guidelines for copyright enforcement, and several jurisdictions
have updated their laws to align with these standards.
India should participate in global intellectual property discussions and modernize its copyright framework to align with WIPO’s evolving principles, ensuring international compatibility in copyright enforcement.
5.2.vi. Public
Awareness and Digital Literacy Programs
A significant challenge in copyright
enforcement is the lack of public awareness regarding intellectual
property rights and the consequences of copyright infringement.
India should introduce educational initiatives that promote awareness
about copyright laws among digital content creators, consumers, and businesses,
ensuring ethical content use and compliance with legal norms.
By adopting these measures,
policymakers and legal experts can ensure a balanced and effective copyright
system that supports intellectual property rights while fostering
accessibility, legal compliance, and technological advancement.
[1] What
is AI? / Basic Questions. (n.d.). Professor John McCarthy. http://jmc.stanford.edu/artificial-intelligence/what-is-ai/
[2] Nystrom, P.
(2017, July 13). Machine Learning: Intro to the Future of Computing.
Global Web and Mobile Application Development Company | Seamgen. https://www.seamgen.com/blog/machine-learning-future-computing
[3] Boyden,
B. E. (2016). Emergent Works. The Columbia Journal of Law & The Arts,
39(3), 377–394. https://doi.org/10.7916/jla.v39i3.2077
[4] Arya, Nisha.
(2024, June 10). Theory of Mind AI in Artificial Intelligence.
EJable. https://www.ejable.com/tech-corner/ai-machine-learning-and-deep-learning/theory-of-mind-ai-in-artificial-intelligence/
[5]
Elizabeth Rocha, "Sophia: Exploring the Ways Al May Change Intellectual
Property, Protections", 28 Dr PAUL J. ARt, TECH. & INTELL. PROP. 1.
126 (2018).
[6] Copyrights act of 1957
[7] Kevin P.
Murphy, "Machine Learning: A Probabilistic Perspective", MIT Press
(2007)
[8] Michael
Copeland, "What's The Difference Between Artificial Intelligence, Machine
Learning, And Deep Learning?", NVIDIA BLOG (July 29, 2016), Available At Https://Blogs.Nvidia.Com/Blog/2016/07/29/Whats
Difference-Artificial- Intelligence-Machine-Learning-Deep-Learning-Ai/ ( Visited On Feb. 14, 2025).
[9] Sujith Ravi,
"On-Device Machine Intelligence", Google Research Blog (Feb. 9, 2017)
Available At Https://Perma.Cc/WQ8L-WS5D (Visited On Feb 13, 2025).
[10] Brendan Mcmahan
& Daniel Ramage, "Federated Learning: Collaborative Machine Learning
Without Centralized Training Data", Google Research Blog Available At
Https://Perma.Cc/XVA2- J96J(Visited On Feb 13, 2025).
[11] Andres
Guadamuz, "Artificial Intelligence And Copyright", 23 WIPO MAG (2017)
Http:/Www.Wipo.Inti Stpo Magazinerenl20l7/05/Article_0003.Html
[Https:/Iperma.Cc/SD4Q-KE9E (Visited On Feb 13, 2025).
[12] R Chesney, DK
Citron, "Deep Fakes: A Looming Challenge For Privacy, Democracy, And
National Security" 107 California Law Review 17 (2019.)
[13] Ravindra Kumar
& Pankaj Kumar, Training AI And Copyright Infringement: Where Does The Law
Stand?, 2 INDIAN J. INTEGRATED RSCH. L. 1 (January-February 2022).
[14] Ibid
[15]
Thushar V. Senan, Abey Augustine & Aswathy Krishnan, AI, Creativity, And
Copyright Law In India: Navigating The Boundaries Of Originality And
Authorship, 6 INT'l J.L. MGMT. & HUMAN. 2941 (2023).
[16] Dr
Mohan Dewan, Vicarious Liability In IP Violations, Lexology (Aug. 30,
2021), Https://Www.Lexology.Com/Library/Detail.Aspx?G=D191f4e5-3c0c-4541-B9b4-3435ae8c6f33.
[17] Ibid
[18]My
Space Inc. Vs Super Cassettes Industries Ltd. On 23 December, 2016
[19]
Christian Louboutin Sas Vs Nakul Bajaj & Ors On 2 November, AIRONLINE 2018
DEL 1962
[20] Supra
Note 10
[21] Ibid
[22] Ibid
[24] Ibid
[25] Supra Note 10
[26] Ibid
[27] Copyright
Directive (2019/790[27]),
[28] Ravindra Kumar
& Pankaj Kumar, Training AI And Copyright Infringement: Where Does The Law
Stand?, 2 INDIAN J. INTEGRATED RSCH. L. 1 (January-February 2022).
[29] Section 107 Of
US Copyrights Act:-Notwithstanding The Provisions Of
Sections 106 And 106A, The Fair Use Of A Copyrighted Work, Including Such Use By Reproduction In
Copies Or Phonorecords Or By Any Other Means Specified By That Section, For Purposes
Such As Criticism, Comment, News Reporting, Teaching (Including
Multiple
Copies For Classroom Use), Scholarship, Or Research, Is
Not An Infringement Of Copyright. In Determining Whether The Use Made Of A Work
In Any Particular Case Is A Fair Use The Factors To Be Considered Shall
Include—
(1) The
Purpose And Character Of The Use, Including
Whether Such Use Is Of A Commercial Nature Or Is For Nonprofit Educational
Purposes;
[30] Authors Guild,
Inc. V. Hathitrust, 755 F.3d 87, 103 (2d Cir. 2014)
[31] NXIVM Corp. V.
Ross Inst., 364 F.3d 471 (2d Cir. 2004)
[32] Oracle America, Inc. V. Google LLC 886 F.3d 1179 (Fed. Cir. 20a18).
[33] Cartoon Network
LP, LLLP V. CSC Holdings, Inc., 536 F.3d 121, 129-30 (2d Cir. 2008)
[34] Burrow-Giles
Lithographic Co. V. Sarony, Ill U.S. 53 (1884).
[35] Sega
Enterprises Ltd. V. Accolade, Inc., 977 F.2d 1510 (9th Cir. 1992)
[36] Kelly V. Arriba
Soft Corp., 336 F.3d 811 (9th Cir. 2003); Perfect 10, Inc. V. Amazon.Com, Inc.,
508 F.3d 1146 (9th Cit. 2007).
[37] Authors Guild
V. Google Inc., 804 F.3d 202 (2d Cir. 2015).
[38] Harper &
Row Publ V. Nation Enters 471 U.S. 539, 560 (1985).
[39] Edward Rex,
"Jukedeck: Building A Creative Al Business", Available At Https://
Entrepreneurship.Blog.Jbs.Cam.Ac.Uk/Author/Ed-Newton-Rex/(Visited On Feb. 14,
2025)
[40] Ibid
[41] Joseph Straus,
"Artificial Intelligence-Challenges And Chances For Europe", 29
European Review (2020)
[42] Ibid
[43] Supra Note 8
[44] Ibid
[45] Ibid
[46] Enrico Bonadio
& Luke Mcdonagh, "Artificial Intelligence As Producer And Consumer Of
Copyright Works: Evaluating The Consequences Of Algorithmic Creativity",
I.P.Q. 112, 13(2020) Https://Eprints.Lse.Ac.Uk/105272/1/Bonadio_Mcdonagh_IPQ_2020.Pdf
[47] Supra Note 8
[48] Sean Flynn Et
Al., Implementing User Rights For Research In The Field Of Artificial
Intelligence: A Call For International Action, 2020 SSRN Elec. J., , Https://Doi.Org/10.2139/Ssrn.3578819.
[49] Supra Note 8
[50] Ibid
[51] Ted Shapiro
& Sunniva Hannson, "The DSM Copyright Directive- EU Copyright Will
Indeed Never Be The Same", EIPR 404, 6 (2019).
Https://Www.Wiggin.Co.Uk/App/Uploads/2019/08/2019_41_EIPR_Issue_7FINAL_SHAPIRO.Pdf
[52] Daniel Gervais,
"Exploring The Interfaces Between Big Data And Intellectual Property
Law", 10 J. INTELL. PROP. INFO. TECH. & ELEC. COM. L. 3 (2019) Https://Papers.Ssrn.Com/Sol3/Papers.Cfm?Abstract_Id=3360344
[53] Ibid
[54] Pierre N.
Leval, "Towards A Fair Use Standard", 103 HARV. L. REV. 1105, 1111
(1990). Https://Archive.Blogs.Harvard.Edu/Copyrightosc/2020/02/26/
[55] Article 7 In Directive 2019/790 (CDSMD)
[56] Article 4 In Directive 2019/790 (CDSMD)
[57] Christophe
Geiger Et Al., Text And Data Mining In The Proposed Copyright Reform:
Making The EU Ready For An Age Of Big Data?, 49 IIC - Int'l
Rev. Intell. Prop. & Competition L. 814, (2018), Https://Doi.Org/10.1007/S40319-018-0722-2
[59] PTI, Central
Task Force On AI Recommends Setting Up Of N-AIM, Indian Express (Mar. 28,
2018), Https://Indianexpress.Com/Article/India/Central-Task-Force-On-Ai-Recommends-Setting-Up-Of-N-Aim-5114130/
.
[60] Ibid
[61] Section 52(1)
Of Indiasn Copyrights Act Of 1957 The
Following Acts Shall Not Constitute An Infringement Of Copyright, Namely,--
1[(A) A Fair Dealing
With Any Work, Not Being A Computer Programme, For The Purpose Of--
(I) Private Or
Personal Use, Including Research;
(Ii) Criticism
Or Review, Whether Of That Work Or Of Any Other Work;
(Iii) The
Reporting Of Current Events And Current Affairs, Including The Reporting Of A
Lecture Delivered In Public;
Explanation.--
The Storing Of Any Work In Any Electronic Medium For The Purposes Mentioned In
This Clause, Including The Incidental Storage Of Any Computer Programme Which
Is Not Itself In Infringing Copy For The Said Purposes, Shall Not Constitute
Infringement Of Copyright.]
[62] Section
52(1)(A) A Fair Dealing With A Literary, Dramatic, Musical Or Artistic Work
Private Use Including Research
[63] Saregama India Ltd. & Ors. V.
Alkesh Gupta & Ors. (2013)
[64] Tips Industries
Ltd. V. Wynk Music Ltd AIRONLINE 2019
BOM 1452
[65] Divij Joshi,
"Crawl Cautiously: Examining The Legal Landscape For Text And Data Mining
In India" Available At Https://Spicyip.Com/2020/06/Crawl-Cautiously-Examining-The-Legal-Landscape-For-Text-And-Data-Mining-In-India-Part-I.Html
Visited On Feb. 13 2025).
Article Information
AI AS A CREATOR: LEGAL CHALLENGES IN COPYRIGHT INFRINGEMENT AND FAIR USE UNDER INDIAN LAW.
Authors: ELEENA ANIL VADAKKEKARA, DR. ARADHANA SATISH NAIR
- Journal IJLRA
- ISSN 2582-6433
- Published 2025/03/15
- Issue 7
About Journal
International Journal for Legal Research and Analysis
- Abbreviation IJLRA
- ISSN 2582-6433
- Access Open Access
- License CC 4.0
All research articles published in International Journal for Legal Research and Analysis are open access and available to read, download and share, subject to proper citation of the original work.
Disclaimer: The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of International Journal for Legal Research and Analysis.