Copyright Issues Raised By Generative Artificial Intelligence (GenAI)

Author: LawSphereX

Category: Technology, Cyber & Privacy

Published: 2026-08-23T18:58:50.928Z

Updated: 2026-08-30T07:11:01.584Z

ARTIFICIAL INTELLIGENCE (GENAI) Mahira Saraf, Vivekananda Institute of Professional Studies, Guru Gobind Singh Indraprastha University, New Delhi

ABSTRACT

This paper explores the intricate relationship between generative artificial intelligence (AI) and traditional AI, highlighting the unique capabilities of generative AI in creating novel content. It delves into the operational mechanisms of generative AI, its diverse applications across various sectors, and the consequent legal challenges, particularly concerning copyright ownership. Focusing on the Indian Copyright Act, 1957, and OpenAI's Terms of Use, the research examines the complexities of attributing authorship and ownership to AI-generated works. It analyses the "Skill and Judgment with flavour of creativity" test for originality and the implications of joint authorship, revealing the inadequacies of existing legal frameworks in addressing the nuances of AI-generated content.

The paper concludes by emphasising the need for legal adaptation to accommodate the rapid advancement of generative AI, advocating for a balanced approach that fosters innovation while protecting creators' rights.

Generative Artificial Intelligence vis-à-vis Artificial Intelligence Artificial Intelligence is a term that most of us have been using recurrently in our day-to-day lives, from relying on the AI Overview provided by Google Gemini to hopping onto the very recent “Studio Ghibli” trend on Instagram. Yet, there is no legally recognized definition of the term. The Father of Artificial Intelligence, John McCarthy, has defined Artificial Intelligence as “The science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable.

”1 Thus, while AI is a technology that can perform tasks requiring human intelligence, Generative AI is an advanced type of AI, which, as per the Organisation for Economic Co-operation and Development (OECD) is "a technology that can create content, including text, images, audio, or video, when prompted by a user. "2 Thus, Generative AI has the potential to develop novel and original content based on the patterns it has learnt in contrast to AI, which resembles a master tactician, capable of making astute decisions within a defined framework of rules. Other classic examples of traditional AI include voice assistants such as Siri or Alexa, recommendation systems like those on Netflix or Amazon, or even Google’s search algorithm.

These AI systems are meticulously trained to operate within specific parameters, excelling in performing designated tasks efficiently, yet lacking the capacity for genuine innovation or original creation. How does Generative AI work? Generative AI models like OpenAI’s ChatGPT and DALL-E, Google’s Gemini, Stability AI’s Stable Diffusion program, Copilot and Cortana for Microsoft users, Anthropic’s Claude models are trained on extensive datasets of existing content, enabling them to generate novel outputs. By discerning intricate patterns within the data through probability distributions, these models, when prompted, synthesise outputs that mirror or align with the identified patterns. Generative AI’s training dataset is based on deep learning, which, essentially, is a result of 30 years of

1 John McCarthy, What is artificial intelligence? The Kurzweil Library

https://www.thekurzweillibrary.com/what-is-artificial-intelligence (last visited on 29 March, 2025)

2 WIPO, Patent Landscape Report - Generative Artificial Intelligence (GenAI)

https://www. wipo. int/web-publications/patent-landscape-report-generative-artificial-intelligence-genai/en/1generative-ai-the-main-concepts. html (last visited on 29 March, 2025) cumulative progress to increase ("deepen") the number of layers of neural networks. 3 These neural networks are modeled after the human brain and have the ability to process and interpret more intricate patterns than conventional machine learning methods. Most significantly, these neural networks operate autonomously, without the need for human oversight, to recognise distinctions and patterns within the training data. 4 Scope of Generative AI The full potential of Generative AI cannot yet be precisely defined, but a general understanding can be derived from its widespread applications across the world today.

To explore these capabilities, there is no better source than ChatGPT itself, which can provide a clear breakdown of the tasks that Generative AI is currently able to perform. Here is a list of tasks ChatGPT believes Generative AI can do-

1. It can assist with content creation, writing articles, blogs, essays, reports, stories, and

poems. It also helps with creative writing, generating fictional narratives, dialogues, or even screenplays, and can condense lengthy texts into concise summaries. It offers translation services between languages, paraphrases text to change wording while maintaining meaning, completes unfinished ideas or sentences, and produces marketing copy for advertisements, emails, or product descriptions. Additionally, it helps with drafting professional or personal emails, legal writing, and even provides research assistance by suggesting outlines or content for academic or professional work.

2. In the domain of code generation, Generative AI writes code in various programming

languages like Python, JavaScript, or HTML, assists in debugging code, generates detailed code documentation, creates automation scripts, and suggests algorithms for solving complex problems.

3. In image generation and editing, it creates original digital artwork from text

descriptions, edits images by adding or removing elements, enhances image quality, 3 James Chen et al. What Is a Neural Network? A neural network is a series of algorithms that endeavors to recognize underlying relationships in a set of data through a process that mimics the way the human brain operates. In this sense, neural networks refer to systems of neurons, either organic or artificial in nature, Investopedia. https://www.investopedia.com/terms/n/neuralnetwork.asp (last visited on 30 March, 2025) 4 Elysse Bell et al. Generative AI: How It Works and Recent Transformative Developments, Investopedia. https://www.investopedia.com/generative-ai-7497939 (last visited on 30 March, 2025) applies artistic style transfers, and helps generate 3D models and designs.

4. For audio and music creation, Generative AI composes music in various genres,

generates sound effects for media production, synthesizes human-like speech from text, enhances the quality of recorded audio, and produces narrations for podcasts or audiobooks.

5. In video generation, it can create short videos or animations from text or images, assist

with video editing tasks such as trimming and merging, produce AI-driven deepfakes by altering faces or voices, and convert scripts into animated videos.

6. Generative AI also serves as a virtual companion and chatbot, engaging in natural

conversations, offering customer support, and functioning as a personal assistant to help with scheduling, reminders, and daily tasks.

7. In design and visual arts, it aids in creating logos, banners, posters, and other visual

content, generates UI/UX designs for websites or mobile apps, suggests fashion designs, and proposes interior design ideas for room layouts and color schemes.

8. In the area of data analysis and insights, it generates synthetic data for testing, creates

data visualizations such as charts and graphs, builds predictive models to forecast trends, and performs statistical analysis on raw datasets.

9. Generative AI can also provide personalization and recommendations, suggesting

products, services, or content based on user preferences, creating personalized educational content, and generating customized fitness and nutrition plans.

10. In simulation and gaming, it creates new game content such as levels, characters, or

narratives, designs virtual environments for immersive experiences, and develops intelligent character dialogue for video games.

11. In the business and marketing domain, Generative AI drafts business proposals,

generates market reports and competitive analyses, creates advertising campaigns, and suggests brand names, taglines, or slogans.

12. It is also useful in scientific research and innovation, helping to draft research papers,

generate hypotheses, aid in drug discovery, and assist with designing new materials for various industries.

13. For language learning and teaching, it provides real-time translations, acts as a tutor for

language learners, and corrects grammar, punctuation, and writing style in text.

14. In the field of social media content, it creates engaging posts, generates memes from

text descriptions, and helps plan content for influencer marketing campaigns.

15. In medical and healthcare, it analyzes medical images to assist with diagnoses, offers

general symptom-based medical advice, tracks health metrics, and drafts summaries of medical data.

16. Generative AI is a valuable tool for education and learning, assisting with lesson

planning, creating interactive quizzes and tests, and tutoring students in various subjects.

17. For legal assistance, it helps draft contracts and agreements, analyze legal cases, and

summarize lengthy legal documents into concise explanations.

18. In the realm of personal growth and coaching, Generative AI helps set personal growth

goals, provides mental health support through personalized mindfulness exercises, and offers life coaching for career planning, productivity, or personal challenges.

19. In news and journalism, it generates news articles based on current events, assists with

fact-checking, and identifies emerging trends in media or public discourse.

20. Finally, in the area of ethical decision-making, it helps evaluate the ethical implications

of AI usage, detects biases in AI models or datasets, and suggests improvements for reducing such biases. The growing prevalence of generative AI tools prompts the question of whether anyone can claim copyright over the content produced by these programs.

Ownership of Copyright The rise of generative AI has introduced complex copyright concerns. As AI systems create original content, questions arise regarding who, if anyone, owns the rights to this material. Copyright is a legal term used to describe the rights that creators have over their literary and artistic works. Works covered by copyright range from books, music, paintings, sculpture, and films, to computer programs, databases, advertisements, maps, and technical drawings. 5 Generative AI, while having the capacity to produce novel works, can also abet criminal activities such as creating deepfakes, voice mimicking, and generating images or videos which can cause injury to religious, moral and personal sentiments of the society at large. In such cases, the question of the ownership of such work plays a pivotal role in the investigation.

Under the Copyright Act, 1957, Copyright vests in all original literary, dramatic, musical and artistic works, cinematograph films, and sound recordings. 6 The word "original" is not defined by the legislature in the USA, UK and India. The courts are often faced with the facts where they have to determine whether work is original or not. The Supreme Court of India laid down the test of “Skill and Judgment with the flavour of creativity” for a work to be original in the case of Eastern Book Company & Ors. v. D. B. Modak & Anr. 7 According to this test, for a work to be original, the author must produce the work with the exercise of their skill and judgment along with a flavour of creativity. Here, the element of creativity need not be very high, but a minimum level of creativity is required.

The concept of originality is inherently tied to the idea-expression dichotomy, as protection is granted only to the original expression of an idea, not the idea itself. This is important to determine the rightful owner of the copyright, i. e. the person who has expressed their original idea. In the case of Najma Heptulla v. M/S Orient Longman Ltd. 8, Justice B. N. Kirpal opined that when two or more people agree to produce a work, wherein one provides the material and the other expresses it in a language which is perceptible to public, then it can be said that there is

5 WIPO, Copyright

6 Section 2(y)

7 (2008) 1 SCC 1.

8 AIR 1989 Del.63

intellectual contribution by those two or more persons, and they are regarded as joint authors or owners. 9 Thus, under the existing framework of Indian Intellectual Property laws, particularly about copyright, extending copyright protection to works created by artificial intelligence (AI) presents significant challenges, as while the Idea of the work may be of the user, it may be expressed by Generative AI. Thus, the issue arises whether the work generated is owned by the GenAI or by the user or is capable of being characterized as a work of joint authorship. Copyright Implications of OpenAI’s Terms of Use The Terms of Use for OpenAI assert that users of the Generative AI retain ownership of all input provided to the database as well as the output generated by the AI, based on the information provided by the user. However, it is difficult to implement such a policy in India.

Section 2(d)(vi) specifies that, for computer-generated works, authorship is attributed to the person who "causes the work to be created. ” Thus, as per the Copyright Act, 1957, OpenAI must also be the owner of the output it generates. If we consider the scope of Assignment of Ownership, Section 18 provides that “The owner of the copyright in an existing work or the prospective owner of the copyright in a future work may assign to any person the copyright either wholly or partially and either generally or subject to limitations and either for the whole term of the copyright or any part thereof”. From this, it can be presumed that for assigning ownership of a work, either wholly or partially, the assignor must be the owner of such work.

As per the “Skill and Judgement with flavour of creativity” test of originality, input given to the GenAI may not necessarily have an element of creativity or skill and judgement. In such a case, the input cannot be assumed to be original and owned by the user, and thus, the output generated by the GenAI cannot be considered as original, thus jeopardizing the claim of OpenAI’s ownership for the Assignment. Even in case the input and output meet the test of Originality, Section 19 requires the assignment to specify the amount of royalties or consideration payable to the author (OpenAI). Additionally, the Terms of Use do not provide for a time period of the assignment of ownership to the user and in such a case, Section 19(5) provides that the deemed period of assignment to be five years, thus resulting in OpenAI to own its generated work after the lapse of this period.

9 1 ALKA CHAWLA, LAW OF COPYRIGHT: COMPARATIVE PERSPECTIVES 41 (1st ed, 2013)

Moreover, the Terms of Use of OpenAI explicitly provides that the work generated by it may not be attributed to it. 10 While this does not result in a violation of the Copyright law in India, as the right of attribution under Section 57 of the Copyright Act is discretionary and can be waived by the author, it poses significant challenges in European Countries like France and Romania. In these jurisdictions, the moral right of attribution is inalienable and cannot be waived, meaning the author must always be credited for their work. 11 Conclusion In conclusion, the emergence of generative AI presents a paradigm shift in content creation, demanding a reevaluation of existing copyright frameworks. While generative AI tools offer unprecedented capabilities, their outputs challenge traditional notions of authorship and originality.

Indian copyright law, with its emphasis on "skill and judgment with a flavour of creativity," struggles to accommodate the nuances of AI-generated works. The ambiguity surrounding authorship, particularly concerning the user's role versus the AI's autonomous creation, complicates the determination of ownership. OpenAI's Terms of Use, which attempt to assign ownership to the user, clash with Section 2(d)(vi) of the Copyright Act, 1957, that attributes authorship to the entity "causing" the work's creation. Furthermore, the complexities of assignments under Sections 18 and 19, including royalty considerations and time limitations, reveal the inadequacy of current legal provisions. The lack of clarity around attribution, though permissible in India, poses significant issues in jurisdictions with inalienable moral rights. The legal landscape must adapt to this technological evolution.

This adaptation requires a nuanced approach, potentially involving legislative amendments to address the specific challenges posed by generative AI. It is imperative to strike a balance between fostering innovation and protecting creators' rights, ensuring that the legal framework remains relevant in the face of rapidly advancing AI technologies. Issues concerning copyright infringement by

10 OpenAI Content Attribution

https://openai.com/brand/

11 Martine Hebette et al. Copyright Law in the EU, EPRS, 174 & 301 (2018)

https://www.europarl.europa.eu/RegData/etudes/STUD/2018/625126/EPRS_STU(2018)625126_EN.pdf (Last visited on 30 March, 2025) GenAI have resulted in the filing of lawsuits, such as by ANI against OpenAI.12 This evolving field necessitates ongoing dialogue among legal experts, technologists, and policymakers to establish clear guidelines and prevent potential exploitation or legal ambiguity.

12 Sakshi Tiwari & Tushar Gaur, The Copyright Conundrum

ANI Vs OpenAI And The Fight For Fair Use, Khurana and Khurana, Mondaq https://www.mondaq.com/india/copyright/1604088/the-copyright-conundrum-ani-vs-openai-and-the-fight-forfair-use (last visited on 30 March, 2025)