Author: Yamini Goel
University: Amity University, Noida
LinkedIn: linkedin.com/in/yamini-goel-34509b415
Abstract
Artificial Intelligence (AI) has transformed the manner in which literary, artistic, musical, and technological works are created. AI-powered tools can now compose songs, generate paintings, write articles, and produce computer codes with minimal human intervention. This technological advancement has posed a fundamental legal question: Who owns the copyright in works generated by Artificial Intelligence? The Indian copyright regime, primarily governed by the Copyright Act, 1957, was enacted at a time when creativity was considered an exclusively human endeavour. Consequently, the Act does not expressly address ownership rights in AI-generated creations.
This article analyses the legal position of AI-generated works in India, examines judicial developments and international practices, and evaluates whether existing copyright principles are sufficient to address the challenges posed by artificial intelligence.
Introduction
Artificial Intelligence refers to computer systems capable of performing tasks that ordinarily require human intelligence, such as learning, reasoning, problem-solving, and content creation. The emergence of generative AI applications such as ChatGPT, Midjourney, and DALL-E has significantly blurred the distinction between human and machine authorship.
Copyright law is premised on the concept of originality and authorship. Traditionally, an author is a natural person who exercises skill, labour, and judgment in creating an original work. However, when an AI system independently creates a painting or writes a novel, identifying the legal author becomes a complex issue.
The absence of specific legislation in India regarding AI-generated works has created uncertainty in relation to ownership, infringement, and enforcement of copyright.
Legal Framework under Indian Copyright Law
The principal legislation governing copyright in India is the Copyright Act, 1957.
Section 13 – Works in Which Copyright Subsists
Copyright subsists in:
1. Original literary works;
2. Dramatic works;
3. Musical works;
4. Artistic works;
5. Cinematograph films; and
6. Sound recordings.
The provision, however, does not expressly include works generated autonomously by artificial intelligence.
Section 2(d) – Meaning of Author
Section 2(d) defines an author differently for different categories of works. Importantly, it states:
In relation to a computer-generated work, the author is “the person who causes the work to be created.”
This provision has become the foundation of the debate concerning AI-generated works. The phrase “person who causes the work to be created” may include:
The programmer who developed the AI system;
The user who provided prompts;
The company that owns the AI software; or
Another person exercising effective control over the creation process.
The Act, however, does not define the extent of human involvement necessary to claim authorship.
The Proof: Why AI Challenges Copyright Law
1. Absence of Human Creativity
Copyright protection is traditionally granted to works involving human intellectual effort. Autonomous AI systems can create content without direct human creativity, making it difficult to satisfy the requirement of originality.
2. Ambiguity of Ownership
Several stakeholders contribute to AI-generated content:
Developers of AI models;
Owners of datasets;
Users giving instructions or prompts;
Companies operating AI platforms.
Determining which of these parties possesses ownership rights remains legally uncertain.
3. Training on Copyrighted Material
Most AI systems are trained on enormous datasets, often containing copyrighted works. Questions arise regarding:
Whether such training amounts to infringement;
Whether outputs substantially reproduce existing works; and
Whether original authors deserve compensation.
4. Lack of Legislative Recognition
Indian copyright law was enacted long before the emergence of modern AI technologies. Consequently, it does not specifically regulate machine-generated creativity.
Judicial and International Developments
1. The “Monkey Selfie” Case
Naruto v. Slater, 888 F.3d 418 (9th Cir. 2018)
Facts
A monkey named Naruto clicked photographs using a photographer’s camera. An animal rights organisationargued that the monkey owned the copyright.
Judgment
The United States Court of Appeals held that animals cannot hold copyright because copyright law recognises only human authors.
Significance
The case reinforced the principle that non-human entities cannot be authors under traditional copyright law. By analogy, AI systems also cannot presently be recognised as copyright owners.
2. Thaler v. Comptroller-General of Patents, Designs and Trade Marks (2023) UKSC 49
Facts
Computer scientist Stephen Thaler sought copyright and patent protection for works generated by an AI system called DABUS.
Judgment
The UK Supreme Court held that an AI system cannot be recognised as an inventor or legal person under existing legislation.
Significance
The decision highlighted the inadequacy of current intellectual property laws in addressing autonomous AI creations.
3. Zarya of the Dawn Copyright Decision (United States Copyright Office, 2023)
Facts
A graphic novel contained images generated through the AI platform Midjourney.
Decision
The U.S. Copyright Office granted protection only to the human-authored portions and denied protection to purely AI-generated images.
Significance
The decision adopted a “human authorship” approach, suggesting that copyright subsists only where substantial human creativity is involved.
4. Eastern Book Company v. D.B. Modak, (2008)
Facts
The issue before the Supreme Court was whether edited judicial reports possessed sufficient originality to attract copyright protection.
Judgment
The Court adopted the doctrine of “modicum of creativity”, holding that a work must involve skill and judgment and not merely mechanical labour.
Significance
This case is relevant to AI-generated works because it emphasises the necessity of human intellectual contribution for copyright protection.
Position of AI-Generated Works in India
The Indian legal position remains uncertain.
A possible interpretation of Section 2(d)(vi) is that the person who causes the work to be created may be considered the author. Depending upon the circumstances, this person could be:
1. The user who enters prompts and exercises creative control;
2. The programmer who designed the AI model; or
3. The company operating the AI system.
However, if the AI autonomously generates content without meaningful human intervention, it becomes difficult to identify an author under existing law.
Need for Legal Reforms
India requires a comprehensive legislative framework addressing AI-generated works. Possible reforms include:
1. Statutory Definition of AI-Generated Works
The Copyright Act should expressly define AI-generated creations.
2. Recognition of Human Contribution
Copyright should subsist only where there issubstantial human creativity and control.
3. Clarification of Ownership Rules
Legislation should specify whether ownership belongs to users, developers, or AI companies.
4. Regulation of Training Data
Clear rules should govern the use of copyrighted material in training AI models.
5. Creation of a Sui Generis Regime
India may consider a separate form of intellectual property protection for autonomous AI creations.
Conclusion
Artificial Intelligence has fundamentally altered the landscape of creativity and authorship. Indian copyright law, however, remains rooted in the assumption that every creative work originates from human intellect. The absence of specific provisions regarding AI-generated content has generated significant uncertainty concerning ownership and protection.
Presently, the most plausible interpretation under Indian law is that the person who causes the AI-generated work to be created may claim authorship, provided there is sufficient human involvement. Nevertheless, fully autonomous creations continue to exist in a legal vacuum.
As artificial intelligence becomes increasingly sophisticated, legislative intervention is imperative. A balanced framework protecting innovation while safeguarding human creativity will determine the future of copyright law in the age of artificial intelligence.
Frequently Asked Questions (FAQs)
Q1. Does Indian law recognise Artificial Intelligence as an author?
No. The Copyright Act, 1957 recognises only legal persons and does not confer authorship upon AI systems.
Q2. Can a person claim copyright over AI-generated content?
Possibly, if the individual exercised substantial creative control and can be considered the person who caused the work to be created.
Q3. Does copyright automatically subsist in purely AI-generated works?
The legal position remains uncertain because Indian legislation does not specifically address autonomous AI creations.
Q4. Can AI training on copyrighted material amount to infringement?
Yes. If copyrighted works are used without authorisation and statutory exceptions do not apply, infringement issues may arise.
Q5. Does India require a separate law for AI-generated works?
Considering the rapid advancement of AI technologies, many scholars support the introduction of a specialisedlegislative framework.
References
1. Copyright Act, 1957.
2. Eastern Book Company v. D.B. Modak, (2008) 1 SCC 1.
3. Naruto v. Slater, 888 F.3d 418 (9th Cir. 2018).
4. Thaler v. Comptroller-General of Patents, Designs and Trade Marks, (2023) UKSC 49.
5. United States Copyright Office, Zarya of the Dawn Decision (2023).
6. WIPO, “Artificial Intelligence and Intellectual Property Policy.”
7. Nandan Kamath, Law Relating to Computers, Internet and E-Commerce.
