Abstract
Artificial intelligence tools are now capable of generating artworks, music , photos and videos that appear real. This development in Artificial Intelligence has made it difficult for copyright laws that protect human work. This article examines the concept of authorship under Section 2(d)(vi) of the Copyright Act, 1957 and analyses how Indian law deals with AI-generated content and protect human work. It focuses mainly on the requirement of human involvement using the Suryast case and important judicial decisions on originality.
I. Introduction
Copyright law aims to encourage creativity by giving authors exclusive rights over their original works. In India, the Copyright Act, 1957 forms the main legal framework. However, this law was designed long before the rise of modern generative AI systems such as ChatGPT, Gemeni, or Grok. These tools can now produce complex works with little direct human effort after the initial prompt.
This article takes a specific perspective: it analyses the human authorship Imperative under Section 2(d)(vi) of the Act. The focus is on whether AI-generated works can qualify for copyright protection and who can be considered the author. The analysis relies on statutory provisions, court judgments, and scholarly writings.
II. The Statutory Position on Authorship
“Author” is defined in Section (d) of The Copyright Act, 1957. In most works, author is the person who creates it. Significantly, the 1994 Amendment included clause (vi), which states that in the case of a computer-generated literary, musical or artistic work, the author is :the person who causes the work to be created.”
This provision was introduced to deal with works produced with the help of computers that existed at that time. Scholars note that it was not drafted with today’s advanced generative AI in mind, which can make independent creative choices based on training data. Section 13 of the Act protects original works, but originality is not defined in the statute and has been explained by courts.
Indian law generally requires a natural person (a human being) as the author. Companies or other entities can own copyright through assignment under Section 18, but the first owner is usually a human. This human-centric approach appears in several court decisions and aligns with the overall scheme of the Act.
III. The Test of Originality – Learning from Eastern Book Company
A key Supreme Court judgment that explains originality is Eastern Book Company v. D.B. Modak and anr. (2008). Before this case, some courts followed the “sweat of the brow” approach, which protected works based only on the effort and labour put into them. The Supreme Court rejected this and also did not fully adopt the stricter American “modicum of creativity” test.
Instead, the Court held that a work must involve the exercise of skill and judgement that shows a “flavour of creativity.” The judgment states that the author must produce something that is not merely the result of labour and capital but involves intellectual effort. Even if the creativity is not highly novel, there must be some substantive variation through selection, coordination, or arrangement.
This standard is important for AI works. If the content is generated using an AI tool almost entirely on its own, then it becomes difficult to prove the required human skill and judgement. On the other hand, if a person carefully drafts prompts, selects outputs, edits them substantially, and arranges the final result, copyright may protect the human contributions.
IV. The Suryast Case and Its Lessons
The most relevant Indian example is the case involving the artwork Suryast. Ankit Sahni used the RAGHAV Artificial Intelligence Painting App to create the image. He provided a base photograph and a style reference. In 2020-2021, the Copyright Office initially granted registration and listed both Sahni and the AI tool RAGHAV as co-authors. This was widely reported as a significant step.
However, the Office later issued a withdrawal notice. It asked for clarification regarding the legal status of the AI system as an author and referred to Sections 2(d)(iii) and 2(d)(vi). The registration was effectively withdrawn, and the Office clarified that a human author is mandatory for copyright claims in India.
This case illustrates the practical difficulty. Even when a human starts the process by giving inputs, if the AI performs the main creative work (such as interpolating styles and generating the final image), the Office appears unwilling to recognise the AI as an author or to grant full protection without sufficient human control. The episode highlights the gap between the wording of Section 2(d)(vi) and the realities of generative AI.
V. Major Challenges in Interpretation
Several issues arise when trying to apply the existing law:
First, what does “causes the work to be created” mean in the AI context? Giving a simple prompt may not be enough, as it is more like giving instructions than creating the expression. Legal writers have pointed out that prompts often convey ideas, which are not protected under copyright. The real expression comes from the AI model.
Second, there is a need to distinguish between AI-assisted works and AI-generated works. In assisted cases, where humans retain substantial creative control through editing and arrangement, protection is more likely. Purely autonomous outputs raise serious doubts.
Third, questions of infringement during AI training remain open. AI models learn from large datasets that often include copyrighted material. Section 52 of the Act provides fair dealing exceptions, but their application to commercial AI training is not yet settled.
Court decisions such as Rupendra Kashyap v. Jiwan publishing House have reinforced that only natural persons can be authors in the true sense. Scholarly articles also support the view that the Act’s structure is built around human creativity.
VI. Comparative Perspective
Other countries have taken different approaches. The United Kingdom has a clearer provision that attributes computer-generated works to the person who made the necessary arrangements. The United States follows a strict human authorship requirement, as seen in decisions of the U.S. Copyright Office. India’s position seems closer to emphasising human involvement while retaining some flexibility through the “person who causes” language.
International trends suggest that many jurisdictions are still grappling with these issues, and India can learn from these developments without losing its own balance between innovation and protection of human authors.
VII. The Way Forward
The current legal framework leaves creators and businesses in a state of uncertainty. The Copyright Office could issue detailed guidelines on disclosure of AI use and the level of human contribution required. In the longer term, Parliament may consider amendments to address generative AI specifically — perhaps by creating limited rights for certain AI outputs or by clarifying the threshold for protection.
Any changes must carefully balance the need to encourage technological progress with the protection of human creativity and the public interest in access to knowledge.
Conclusion
Section 2(d)(vi) of the Copyright Act, 1957 gives a starting point for handling computer-generated works, but it needs careful interpretation in the age of generative AI. Judicial decisions like Eastern Book Company v. D.B. Modak and the handling of Suryast case shows that Indian law still necessitates meaningful human authorship. While the law must evolve with technology, it should not lose sight of the fundamental connection between copyright and human creative effort. Clearer guidelines and possible legislative updates will help India address these challenges effectively.
THIS ARTICLE IS WRITTEN BY MITANSH YADAV FROM ASIAN LAW COLLEGE
REFERENCES
Generative AI & Copyright Law in India: Who Owns the Output?, Lexology (16 July 2025).
Impact of AI-Generated Works on Indian Copyright Law: Authorship and Originality Challenges, International Journal of Law and Legal Research (9 September 2025).
Eastern Book Company v. D.B. Modak (2008) 1 SCC 1.
The impact of generative AI on copyright as Indian jurisprudence evolves, IAM Media (25 November 2025).
VK Ahuja, Artificial Intelligence and Copyright: Issues and Challenge, Indian Law Institute Journal.
Rupendra Kashyap v. Jiwan Publishing House Pvt. Ltd., Delhi High Court (relevant observations on authorship).
Ayushee Sinha, AI-Generated Works in IP Laws: Contemporary Authorship and Ownership Challenges, CNLU Journal (2026).
Ankit Sahni’s Suryast registration details as reported in Managing Intellectual Property (2021) and subsequent withdrawal notices.
