Artificial Intelligence and Human Authorship along with the Role of the Copyright Act,1957

Introduction

In recent years, Artificial Intelligence (AI) has emerged as one of the most transformative technologies, heavily influencing our day-to-day lives. From assisting with problem-solving and decision-making to generating text, images, and videos, AI has expanded the boundaries of creativity and innovation. As these technologies continue to evolve, they have challenged the role of human authorship under copyright law.¹

Traditionally, copyright law has continuously protected creative works originating from human intellect and creativity. However, the increasing use of AI has challenged that very notion. If a work is generated with the help of AI and little to no human intervention, where does it fall under the Copyright Act, 1957?

It is crucial to note that AI does not possess consciousness; instead, it generates outputs by analysing vast amounts of training data, identifying statistical patterns, and responding to users’ prompts. This raises concerns for human creators and creates uncertainty around ownership, authorship, and copyright protection.

This article examines whether the requirement of human authorship under the Copyright Act, 1957, is sufficient. It reviews the existing legal framework, discusses the line between AI generation and human authorship, and considers relevant laws outside India through comparative study and case examples.

Legal Framework

The Copyright Act, 1957

The Copyright Act, 1957 defines copyright as a bundle of exclusive economic rights granted to the owner of a protected work. These rights include the authority to reproduce, publish, distribute, communicate, adapt, translate, commercially exploit, or otherwise authorise the use of the work. The scope of these rights depends on the category of work involved. The Act was established to protect humans’ original work, including literature, art, music, and films.² At first, the concept of copyright extended only to books;³ gradually, over time, changes were made to incorporate new technologies and inventions in order to protect the rights, creativity, and originality of human work. Since AI is a comparatively recent technology, the uncertainty regarding authorship under Section 2(d) sits in tension with Section 14 of the Act, as it remains unclear who holds legal entitlement and rights over an AI-assisted work.

Section 2(d) of the Copyright Act, 1957 defines the term “author” by assigning authorship according to the nature of the work. It recognises the writer of a literary or dramatic work, the composer of a musical work, an artist, a photographer, and so on. For each category, the Act designates the author as the person who causes the work to be created.⁴

Section 13 of the Copyright Act specifies the categories of work eligible for copyright protection in India. This section clearly recognises the categories of work described above and the corresponding class of author; however, AI-generated work still sits in a zone of uncertainty. An AI system may well generate output that fits the definitions of a film, a piece of music, a text, or a painting — but the legal issue lies in the statutory requirement of human authorship under the Copyright Act, 1957.

Section 17 of the Copyright Act, 1957 establishes the general rule that the author of a work is the first owner of its copyright. The Act also recognises several exceptions where ownership vests in another person, such as an employer, a person commissioning certain works, the Government, public undertakings, or international organisations, unless otherwise agreed by contract. The principle embodied in Section 17 reinforces the central role of authorship within the Indian copyright regime, since ownership derives its identity from the author. Because the authorship of AI-generated work remains uncertain, determining the first owner of copyright in such a work becomes a genuinely difficult task.

This uncertainty deepens as the question of ownership grows more complex. Ownership could plausibly be assigned to the prompt provider, the developer, or the trainer of the AI system — or to none of them at all. Section 17, however, only provides specific exceptions for works created during employment, commissioned works, and government works; none of these provisions address ownership of works generated through autonomous or largely autonomous AI systems. The Copyright Act, 1957 still fails to provide any clear authority or ownership framework for AI-generated work.

Eastern Book Co. v. D.B. Modak, (2008) 1 SCC 1⁵

The Supreme Court ruled that originality requires a minimal degree of creativity rather than mere labour or effort. This standard implicitly assumes that creative expression originates from a human mind. As noted above, AI generates outputs by analysing vast amounts of training data, identifying statistical patterns, and responding to users’ prompts — meaning that whatever “creative expression” appears in the output must ultimately trace back to that training data.

This, once again, raises the question of ownership and human authorship. It is genuinely difficult to draw firm conclusions here, since the underlying technology is new and has evolved rapidly, challenging lawmakers and the public alike.

Comparative Analysis

Having examined the Indian statutory position, it is useful to compare how other major jurisdictions have approached the same question.

The United Kingdom has adopted a statutory framework that attributes authorship of computer-generated works to the person responsible for making the arrangements necessary for their creation. Although this approach provides a degree of legislative certainty, the relevant provision was enacted well before the emergence of generative AI and offers little clarity where multiple stakeholders — users, developers, and AI providers — each contribute to the creative process.

The United States maintains a strict human-authorship requirement, recognising copyright protection only for works originating from human creative effort. This approach preserves traditional copyright principles but offers limited scope for protecting works produced through substantial human interaction with AI systems.

The European Union follows a comparatively flexible approach, assessing whether a work constitutes “the author’s own intellectual creation” rather than focusing solely on whether AI was involved. The emphasis is placed on the extent of human intellectual contribution and the exercise of creative judgment in producing the final work.

India, by contrast, remains in a state of legal uncertainty. The Copyright Act, 1957 acknowledges computer-generated works by recognising the person who causes the work to be created as its author, but the Act does not address AI-generated work specifically. Unlike the UK, the US, and the EU, India has not yet judicially interpreted this provision in the AI context. This legislative gap has produced real uncertainty around authorship, ownership, and the availability of copyright protection for AI-related work in India.

Case Study: AI Art and Artistic Style

A few months ago, a trend spread across social media in which people uploaded their photographs to AI chatbot apps — Meta AI, Gemini, and ChatGPT among them — and used prompts to have their pictures rendered in the artstyle of Studio Ghibli. Studio Ghibli Inc. is a Japanese animation studio based in Koganei, Tokyo. Hayao Miyazaki, one of the studio’s founding directors, expressed sharp disappointment and disagreement with this use of his work. He noted that his hand-drawn art and animation took great time and effort to produce, and that animators are often overworked and underpaid. This sparked wider concern about the relationship between AI-generated work and human authorship. The resulting images, though AI-generated, are processed and produced through an assessment of existing human-made work — which raises a further question: what happens when AI-generated work closely mimics a distinctive artistic style?

Establishing copyright ownership over a particular artistic style is inherently difficult. Even so, AI-generated work that closely resembles a specific art style raises real legal questions about copyright infringement. If the underlying training data was not licensed, AI tools may have used freely accessible data to train their models — but the process of extracting information from websites or databases can amount to unauthorised copying and storage of copyrighted content, which could itself constitute infringement. It may, however, be contended that, in the absence of statutory protection for ideas under the intellectual property framework, mimicking an artistic style alone does not constitute infringement. Where AI-generated works are presented in a manner likely to cause consumer confusion about their source, official affiliation, or endorsement, this could give rise to a claim for passing off. A claim for unauthorised commercial use of a distinctive style to trade on the source of the artwork can also be made, akin to a claim for unfair competition.⁶

The Raghav (Suryast) Matter

The Suryast matter is one of the earliest instances in India to indirectly raise the copyright implications of AI-generated works. In 2020, Ankit Sahni created an artwork titled Suryast using the AI-based painting application RAAGHAV, and obtained copyright registration from the Indian Copyright Office, listing both himself and the AI system as co-authors. That registration was later questioned by the Copyright Office, exposing the legal uncertainty surrounding the recognition of AI-generated works under Indian copyright law.

The controversy centred on the interpretation of Section 2(d)(vi) of the Copyright Act, 1957, which identifies the author of a computer-generated work as “the person who causes the work to be created.” While this provision acknowledges computer-based works generally, it does not clarify how much human creative involvement is required where an AI system contributes significantly to the final expression. The Suryast matter therefore highlighted the ambiguity surrounding authorship, originality, and ownership of AI-generated works under the existing statutory framework.

The same artwork was also submitted for copyright registration before the U.S. Copyright Office, which refused registration. Sahni argued that his creative choices — in selecting the images, artistic style, and style-transfer parameters — constituted sufficient human authorship. The U.S. Copyright Office Review Board disagreed, holding that the expressive elements of the final work were, in fact, generated by the AI system, and that Sahni had not exercised sufficient creative control over the final output. As a result, the work failed to satisfy the human-authorship requirement for copyright protection in the United States.

The Suryast matter illustrates the growing insufficiency of traditional copyright principles when applied to generative AI. While it does not constitute binding judicial precedent in India, it highlights the interpretive challenges posed by Section 2(d)(vi) and reinforces the need for legislative or judicial clarification of the threshold of human creativity required for AI-generated works to qualify for protection.⁷

Proposed Solution for India

Rather than granting copyright protection to all AI-generated works, or denying it altogether, India should adopt a human-centric approach that recognises copyright only where there is substantial human creative contribution. Parliament should amend Section 2(d) of the Copyright Act, 1957 to clearly define authorship in the context of AI-assisted work, and to clarify the threshold of human involvement required to qualify for protection.

The Government should issue detailed guidelines identifying relevant factors — such as the originality of prompts, the creative choices made during the generation process, the selection and arrangement of AI outputs, and any post-generation modifications. These factors would help courts and the Copyright Office determine whether sufficient human creativity exists in a given case. Such a framework would preserve the fundamental objective of copyright law — protecting human intellectual creativity — while providing legal certainty for artists, developers, businesses, and users who increasingly rely on artificial intelligence.

Conclusion

The future of copyright law should not revolve around recognising artificial intelligence as an author, but around identifying the degree of human intellectual contribution involved in the creative process and setting clear rules around it. Copyright has, since its establishment, rewarded human creativity, and extending that same protection to AI outright would be unfair to the people who put in the time and effort that copyright law was designed to reward. That does not mean AI should be excluded altogether — there should be a clear statutory framework for AI-assisted creations, or for the role AI plays in the creative process, whether as a tool, a co-author, or something requiring citation. As always, a balanced approach is needed, one that allows technological advancement and copyright protection to evolve together rather than in conflict.

Reference(S):

1. Copyright Act, 1957, s. 2(d).
2. Copyright Act, 1957, s. 13.
3. Copyright Act, 1957, s. 17.
4. Eastern Book Co. v. D.B. Modak, (2008) 1 SCC 1 (India).
5. Shardul Amarchand Mangaldas & Co, AI and Human Authorship Requirement: Reflections from the US Position and Implications for Indian Copyright Law.
6. Aashirwad Tripathi, Creative Input Required from an Author While Creating AI-Generated Content to Claim Copyright (2026).
7. Mackenzie Caldwell, “Author”? — Copyright Authorship of AI Art Through a Philosophical Lens, 61(2) (2023).
8. Hafiz Gaffar & Saleh Albarashdi, Copyright Protection for AI-Generated Works: Exploring Originality and Ownership in a Digital Landscape (2024).
9. Thanushkar B & Sailesh Kumar N, Redefining Human Authorship in the Age of Generative AI: An Analysis under the Copyright Act, 1957, VIII(I) (ISSN: 2582-8878).
10. Shardul Amarchand Mangaldas & Co, Navigating AI-Generated Art and Copyright Law (2025).
11. Indian Kanoon, indiankanoon.org.
12. Organisation for Economic Co-operation and Development, oecd.org.