Is a Prompt Enough to Determine Authorship? Reconsidering Copyright Authorship in AI Generated Work

1. Introduction

The World is in the Era of AI, with AI being used for a plethora of work including generative AI for creative and artistic work, which eventually leads to the question “Whether an artistic work created with the assistance of AI can satisfy the requirements for  copyright protection?”, which leads to another question “ Who should be regarded as the possible Author for such artistic work, the AI System itself or the person who provides the Prompt?”

These questions were a part of debate once but now have become a growing challenge for copyright law in different jurisdictions. India, being one of such jurisdictions, has recently faced the challenge while deciding the authorship application for an AI system ‘DABUS’. The application concerned an artwork generated by DABUS(an Generative AI system), where the AI system being identified as the author of the artwork. The Copyright Office rejected the application for recognition of the AI system as an author under the Indian Copyright Law, bringing back focus on the essential requirement of human authorship.

This article examines the existing Indian copyright framework for its adequacy in addressing authorship of generative AI outputs , along with comparative approaches of the United States, United Kingdom and China related to AI generated artwork and its authorship.

The article argues that while AI should not be recognised as an independent legal author under the existing legal framework, AI Generated artwork should not automatically be denied copyright protection. A distinction should therefore be drawn  between work resulting from human creativity and work that is merely the output of generative AI.

2. Traditional concept of Copyright Authorship in India

The copyright protection under the Indian copyright law has traditionally revolved around the protection of an original work which is a result of Intellectual Human creativity for which the author can be identified to be a natural or juristic person. The copyright under Indian law involves  different  concepts such as authorship, originality and ownership which form the foundation for determining who is entitled to protection and under what circumstances. Artificial Intelligence has the ability to produce various forms of work comprising images, text, music, art or any other creative outputs that challenge the traditional concept of copyright as it can easily satisfy the traditional requirement for copyright protection. Thus the concept of authorship, originality and ownership are crucial in regards to AI generated work to determine whom should such right and protection be attributed.

2.1 What is Copyright ?

 Copyright is a statutory intellectual property right protecting original expression in literary, dramatic, musical and artistic works, subject to the requirements of the Copyright Act, 1957. It provides the author with exclusive rights, including reproduction, adaptation, publication and communication of the work.

2.2 Authorship

Under Indian law, the author of different categories of copyright works is identified according to the nature of the work, including:

• Author of the work for the literary or dramatic work
• Composer for musical work
• Artist for artistic work
• Person who has taken the photograph in case of photographs
• Producer in case of cinematograph film, or sound recording
• The Person who causes the work to be created for any computer generated literary, dramatic, musical or artistic work^[1]

2.4 Originality

Section 13 of The Act gives the protection of Original work in areas of literary, dramatic, musical, artistic, cinematograph films and sound recordings^[2]. Although the act does not define originality, the meaning of it has been developed through judicial interpretation.

Originality has been interpreted as independent creation of an author’s expression of idea with active input in the creation of such work with minimal applicable labour, skill and judgement.

2.5 Ownership

It is often confused that ownership and authorship are similar concepts but as per the Indian law, these two are different concepts. Ownership of a copyright work determines who is legally entitled for having rights over the copyrighted work. The author can be the original creator of the work while the owner can be a person other than the one who created the work, like in case of work created under employment or commissioned works, the owner can be the employer and not the employee who created the work.^[3]

Section 17 of the copyright act provides different rules for the ownership of a copyright work and various exceptions for the same, where in general sense the author of the work is determined as the first owner but it may vary in certain circumstances.

3. DABUS Decision: Can an AI be the author under the Indian Law?

An application was made to the Indian Copyright Office seeking registration of an artwork created by an AI system (DABUS) , identifying DABUS as the author^[4], which was eventually rejected by the office on the ground that an AI system cannot be recognised as an author under the Copyright Act, 1957 as it is neither a natural person nor a juristic person recognised by law.

3.1 Facts :

The case involves an artwork “A Recent Entrance to Paradise” which was created through an AI system known as “Device for the Autonomous Bootstrapping of Unified Sentience” or ‘DABUS’. Dr. Stephen L. Thaler made an application for the registration of the work and identified DABUS as its author while claiming himself for the ownership of the copyright. Upon examination of the application, the Office questioned the legal capacity of an AI system to be recognised as an author under the Indian copyright law.

3.2 Decision and Reasoning :

The Copyright Office found that the work satisfies the  requirement of originality under section 13 of the Copyright Act,1957^[5]. Thus, the fact that the work was produced through an AI did not, by itself, make the work unoriginal. However, originality alone was insufficient for registration where the claimed author lacked the legal personality. The office held that DABUS was neither a natural person nor a juristic person recognised under the Indian law and therefore could not be entered as the author of a computer generated work under section 2(d)(vi) of the act.

A further difficulty arose from the claim that DABUS was the author while Dr Thaler was the copyright owner. As DABUS lacked legal personality, it could neither hold nor assign copyright. Treating DABUS as author while attributing ownership to Dr Thaler therefore created a legal inconsistency.

 The Office further considered the distinction between autonomous execution and human conception. Although DABUS performed the computational process that resulted in the artwork, the material before the Office indicated that Dr Thaler had conceived and configured the system and initiated the process that resulted in the particular work. The Office therefore did not treat the system’s autonomous operation as equivalent to legal authorship.

Finally, the Copyright Office emphasised that extending legal personhood or statutory authorship to autonomous artificial intelligence would involve a policy choice for Parliament rather than an administrative reinterpretation of the existing legislation^[6]. The application was therefore rejected in the form in which it was maintained. Importantly, the decision did not establish that every work produced through AI or computational processes is incapable of copyright protection; rather, it distinguished the question of originality of the work from the question of legal authorship of the work.

3.3 Analysis :

The DABUS decision demonstrates that originality and authorship are two distinct legal questions. The fact that an AI-generated work may satisfy the originality requirement does not necessarily mean that the AI system responsible for generating it can itself be recognised as the author. This creates a significant gap in the traditional copyright framework, which was developed on the assumption that the creator of an original work would be a legally recognisable person. As AI systems become increasingly capable of generating creative outputs with limited human intervention, the law must determine whether existing concepts of authorship are sufficient to address such works.  The decision also brings into focus the difficulty of distinguishing between AI as a mere technological tool and AI as an autonomous creator. However, the position becomes less clear where the AI system makes substantial creative decisions and the human contribution is limited to designing, training or initiating the system. The DABUS decision therefore leaves an important question for future cases: “At what point does human involvement become insufficient to justify attributing authorship to the human, and how should the law treat the resulting work when the AI itself cannot be recognised as an author?”

4. Comparative Approach

The difficulty of determining authorship in AI-generated works is not unique to India. The United States, United Kingdom and China adopt different approaches to human authorship and creative contribution, providing a useful basis for assessing the adequacy of the Indian framework.

4.1 United States :

 The position was directly illustrated by the DABUS dispute involving Dr Stephen L. Thaler and the artwork A Recent Entrance to Paradise. Thaler sought copyright registration for the artwork while identifying the AI system DABUS as its creator. The U.S.Copyright Office refused registration on the ground that copyright protection requires human authorship. However, the U.S. position does not mean that the use of AI automatically prevents copyright protection. The U.S. Copyright Office has clarified that works containing AI-generated material may be protected where a human author contributes sufficient expressive elements, such as through creative selection, arrangement or modification. Mere provision of prompts, by itself, is generally insufficient to establish the required human authorship^[7]. While an AI system cannot itself be recognised as the author, human creative contribution can provide the basis for copyright protection.

4.2 United Kingdom :

The United Kingdom adopts a distinctive approach through Section 9(3) of the Copyright, Designs and Patents Act 1988, which expressly provides for “computer-generated works”. Where a literary, dramatic, musical or artistic work is generated by a computer in circumstances in which there is no human author, the law treats the person by whom the arrangements necessary for the creation of the work are undertaken as the author. Such protection is subject to a shorter term of 50 years^[8].

This approach differs significantly from the United States because it does not require the computer itself to possess legal personality or authorship. Instead, the law identifies a human person responsible for the necessary arrangements leading to the creation of the work. In the context of generative AI, this provision may potentially attribute authorship to the person responsible for initiating or arranging the creative process, although its application to modern generative AI raises questions regarding the degree of human involvement necessary. The UK Government has itself acknowledged uncertainty surrounding the continued application and interpretation of the computer-generated works provision in the context of modern AI^[9].

4.3 China :

China provides a comparatively flexible approach to AI-assisted creative works. In Li v. Liu^[10], decided by the Beijing Internet Court in 2023, the claimant created an image using the AI image-generation software Stable Diffusion and subsequently brought a copyright infringement claim when the defendant used the image without permission. The Court recognised the image as a copyrightable artistic work and held that the claimant was entitled to copyright protection.

The Court’s reasoning focused not merely on the fact that AI was used, but on the extent of the claimant’s creative involvement. The claimant had repeatedly adjusted prompts and parameters and made choices concerning the desired appearance of the image. These actions demonstrated personal aesthetic choices and intellectual judgment rather than merely issuing a single basic instruction and accepting the first output.

This approach is significant because it places emphasis on human creative contribution rather than the mere presence of AI in the production process. The Chinese Copyright Law generally recognises natural persons as authors, while also providing for certain circumstances involving legal persons and unincorporated organisations.  Thus, the AI system itself was not treated as the author; rather, copyright protection was linked to the human user’s contribution to the creation of the work.

4.4 Comparative Analysis :

 The comparison shows three approaches: the United States requires human authorship but protects sufficient human expressive contribution; the United Kingdom expressly provides for computer-generated works; and China focuses on meaningful human creative involvement. India, through DABUS, rejects AI authorship but leaves the threshold of human contribution in AI-assisted works uncertain.

5. Legal Gap and Problem in India

The DABUS decision clarifies that artificial intelligence cannot presently be recognised as an author under the Indian Copyright Act, 1957. However, the decision also exposes a significant gap in the existing framework: Indian copyright law does not provide a sufficiently clear test for determining authorship in AI-assisted works where the degree of human involvement varies. Treating all these situations alike may either deny protection to works involving genuine human creativity or grant protection where human contribution is merely mechanical. The comparative approaches demonstrate that other jurisdictions have begun addressing this distinction through human-authorship requirements and statutory provisions for computer-generated works. The absence of a clear statutory or administrative framework in India therefore creates uncertainty regarding authorship, ownership and the threshold of human contribution required for copyright protection. While DABUS establishes that AI itself cannot presently hold authorship, it leaves open the broader question of ‘how Indian law should treat the growing category of AI-assisted creative works’.

6. Conclusion

The emergence of AI-generated creative works has exposed important limitations in the traditional concept of copyright authorship in India. The DABUS decision establishes that AI cannot presently be recognised as an author under the existing statutory framework, while comparative approaches demonstrate different ways of addressing human involvement in AI-assisted creation. The central difficulty for India therefore lies not merely in determining whether AI can be an author, but in identifying when human contribution to an AI-generated work is sufficient to establish authorship and ownership. As AI continues to participate increasingly in creative processes, greater clarity in the Indian copyright framework will be necessary to balance the protection of genuine human creativity with the realities of technological development.

7. Bibliography

Legislation :

1. The Copyright Act, 1957 (India).
2. Copyright, Designs and Patents Act 1988 (UK).
3. Copyright Law of the People’s Republic of China 2020.

Cases / Decisions :

1. Thaler v. Perlmutter, No. 23-5233, 130 F.4th 1039 (D.C. Cir. 2025).
2. Li v. Liu, Jing 0491 Min Chu No. 11279 (Beijing Internet Ct. 2023).
3. In re Dr Stephen L Thaler, Diary No 9356/2022-CO/A, Registrar of Copyrights, Copyright Office, India, Order dated 31 August 2026.
4. Eastern Book Co. v. D.B. Modak, (2008) 1 SCC 1.

Official / Secondary Sources :

1. US Copyright Office, Copyright and Artificial Intelligence (2025).
2. UK Government, Report on Copyright and Artificial Intelligence.
3. Beijing Internet Court, Li v Liu — English translation/materials on AI-generated works.
4. SCC Online, “Original, yes. Author, no : Copyright Office rejects AI authorship, but not AI generated art” (Sept. 1, 2026)
5. Amaal Sheikh, “AI Can Create a Work, but Can it Be its Author? Why the Copyright Office Drew the Line”, The Indian Express (Sept. 2, 2026).
6. LawBhoomi,“Concept of Originality and Subject Matter in Copyright” (February 12, 2026)

^[1] Section 2(d) of Copyright Act,1957

^[2] Section 13 of the Copyright Act, 1957

^[3] Section 17 of the copyright act, 1957

^[4] Application for Registration of Copyright

^[5] Eastern Book Co. v. D.B. Modak, (2008) 1 SCC 1

^[6] Indian Copyright Office

^[7] US Copyright Office

^[8] UK Government, Copyright and Artificial Intelligence

^[9] UK Government, Copyright and Artificial Intelligence

^[10] Li v. Liu, Jing 0491 Min Chu No. 11279 (Beijing Internet Ct. 2023)