1. Introduction: Rethinking Ownership in the Age of AI
Traditionally, it has been a grundnorm that only humans produce novel and creative works that can be protected under the realm of intellectual property law. Contemporarily, generative AI has become magic, and it has led to the prompted examination of AI’s potential for intellectual property protection. Validating IP rights of work created by AI systems working independently without human input or intervention is a tremendous challenge, additionally, there is little judicial or legislative clarity about their protection[1]. Salmond’s expression “person”, “any being to whom the law regards as capable of rights and duties. Any being that is so capable, is a person… even though he is not a man”[2]. Thus, it may be said that the philosophical concept of ‘person’ may encompass both natural and artificial persons and does not only refer to certain biological and psychological traits of humans.
The Indian Copyright Act was amended in 1994 to accommodate claims for works generated by computer; section 2(d)(vi) defines “author” as “in relation to any literary, dramatic, musical or artistic work that is computer-generated, the person who causes the work to be created”[3]. Nevertheless, the protection of works generated by AI remains uncertain due to a narrow interpretation of the word “person”. In the prominent case of Rupendra Kashyap v. Jiwan Publishing House Pvt. Ltd.[4], the High Court of Delhi reinforced the principle that authorship under the Indian copyright regime can only be claimed by natural persons while rejecting the copyright title of an artificial person. The prevalence of AI-generated work has led to legal uncertainties related to its production, protection, and preservation. The current law is on the verge of being outdated, and without any reform, it may hinder innovations.
2. The Unresolved Dilemma: Who Owns AI-Generated Content?
Defining originality and authorship has become one of the most important discourses in determining the applicability of IP rights to AI[5]. Originality is the paramount condition for granting IP rights to any work; however, its degree varies with the jurisdiction. Until 2008, India followed the “sweat of the brow” doctrine that prevails in the UK where the usage of skill and labour of authors is sufficient for granting IP protection, even if creativity is limited[6]. However, in the case of D.B. Modak and Anr. v. Eastern Book Company and Ors.[7], the Supreme Court announced a shift to the “modicum of creativity” which is primarily followed in the U.S. where “independent creation” and “minimum degree of creativity” are a prerequisite for a work to be “original”[8]. However, the applicability of these doctrines to AI-generated works is complex, as AI systems create content by analysing and rearranging vast databases of pre-existing human-created material, frequently without direct human creative input. This demands the question of whether algorithmic results can legally be deemed original, and if so, to whom should authorship be attributed?
Identifying thresholds for human involvement is crucial. Courts of law have progressively differentiated AI-autonomous work from AI-assisted work where humans make artistic decisions, such as enhancing prompts or modifying output[9]. In the landmark case of Thaler v. USPTO[10], it was held that mere launch of an AI process shall be inadequate to grant authorship, but continual collaboration between AI and human, i.e. evaluating and altering outputs, may be eligible. India’s 2025 draft CRI Guidelines suggest a “creative control” test, in which ownership is based on the human’s capacity to direct the AI’s output toward a defined innovative objective[11].
Indian legislations state that IP rights can solely be granted to humans. Therefore, AI, being a machine and not a legal person, is disqualified from owning IP rights or protections[12]. The Hon’ble Supreme Court in the case of Som Prakash Rekhi v. Union of India & Anr[13]concluded that in the eyes of law, ‘personality’ simply means a jurisdictional person who has some inherent rights of suing or being sued by another entity. Therefore, AI cannot be considered a ‘personality’ as it lacks the capability to exercise rights or perform duties independently. The approach becomes unjust and problematic when the ownership is granted to the person who controls AI, ordinarily the developer or user who provides input, in situations when AI independently performs creative tasks. This attributes ownership to those who are not directly involved in the creation of the work.
Section 2(d)(vi) of the Indian Copyright Act clearly states that authorship of computer-generated works is granted to the person who causes their creation[14]. However, this does not imply that the ‘prompt-giver’ causes the generation of work. India follows the modicum of creativity doctrine, where entering a single-line prompt is not enough to grant IP authorship or protection[15], although other jurisdictions that follow the ‘sweat of the brow’ doctrine may grant the prompt-giver the authorship over AI-generated works. In Zarya of the dawn case, the US copyright office granted IP protection to a graphic novel even when it contained images generated using AI systems. The findings were based on the review of the skill, judgment, and creativity used by the author in sequencing, choosing, and modifying the AI-generated images, which eventually altered them into a graphic novel[16]. Subsequently, a policy statement was issued by the USA copyright office stating that a work incorporating AI-generated content may be given copyright if it has enough human authorship to fulfill the criterion for copyright protection.
Ultimately, the dispute revolves around the distinction between human and machine creativity, and how much human input is necessary for a work to be considered original and eligible for copyright.
3. Global Legal Perspectives on AI-Generated IP
- European Union:
The influence of AI on intellectual property has been actively explored by the EU. In 2020, a report[17] published by the European Commission stated that AI should not be granted rights such as patentability or authorship. However, the EU has underlined the importance of adapting existing IP rights frameworks to better suit the changing technology landscape. The EU has also launched negotiations about the creation of new legal categories for AI-generated inventions, particularly in terms of patent law. Furthermore, the new EU AI Act and GDPR provide new data protection and transparency rules for AI development and its relationship with IP rights in Europe. In a recent decision, the Municipal Court of Prague concluded that an image made by AI could not constitute a work capable of copyright protection since it is not the outcome of creative activity of a real person under the Czech Copyright Act[18].
- United States:
Generally, in the U.S, IP rights protect human-created works, but it is silent upon the AI-generated works. In the case of Thaler vs. Vidal case[19] it was established that AI systems cannot be designated as an inventor, but under the Patent Laws, i.e. the inventions created by AI systems have been, in many cases, awarded Patent rights though AI as a system cannot be designated as an inventor. The Berne Convention, 1886[20], and the U.S. Copyright Act, 1976[21], stressed the principle that originality and creative intent should mandatorily come from human mind and laid down that only human-executed works would qualify for IP protection. In 2023, the U.S Copyright Office ruled that since AI systems lack human creativity and originality they cannot be designated as an “author” and as a result, the office rejected copyright protection for AI-generated artwork[22]. In order to structure AI developments as safe, secure and ethical some laws such as the National AI Initiative Act[23] and the Future of AI Innovation Act[24] have been established.
- WIPO
According to a study published by the World Intellectual Property Organization in 2019, nearly 340,000 AI-generated inventions have been patented worldwide[25]. In 2018, Francis Gurry, former Director General of WIPO, stated: “From a purely economic perspective, if we set aside other aims of the IP system, such as ‘just reward’ and moral rights, there is no reason why we shouldn’t use IP to reward AI-generated inventions or creations. But this still requires some thought”[26]. WIPO has taken a leading role in resolving ambiguities around works generated by AI and the ongoing “conversation on AI and IP” has considered a sui-generis system to address complexities by creating a tailored limited protection, mandatory registration and well-defined rights and limitations. This system could boost innovation in AI while acknowledging the value of human’s creations. However, these discussions lack global consensus[27]. WIPO aims at developing harmonised principles or new legal categories that address AI generated works.
4. Rethinking IP: Ownership Dilemmas and Policy Pathways in AI
Determination of ownership of AI-generated work remains one of the most prominent yet contentious challenges. Several discourses suggest distinct models: granting ownership to the organisation operating the AI system as a legal entity; to the developer for creation of the AI system; to the user for giving prompts or piloting output; or consigning AI-generated works to the public domain, implying no exclusive right[28]. Experts suggest granting AI legal rights especially the Intellectual Property rights due to its usefulness in the current world. But the European Parliament’s 2025 draft AI code has rejected this notion saying that the humans or legal entities should only be entitled to rights in order to avoid legal and ethical problems[29].
A sui generis framework granting limited-period rights to AI-created works, incidental to registration and keeping it transparent about AI’s involvement can be a better option[30]. A similar pilot program was introduced by Japan’s 2025 AI Innovation Act where it was made mandatory for developers to share AI’s training data and human supervision[31]. Scholarly recommendations suggest that the user’s ability to select and combine the prompts or resources should play an important role while determining the originality of work. The view was justified by referring to the importance given to curation and arrangement, which is considered as the main creative act during the traditional process of collage-making.
Further, situations where multiple parties are involved in the creation process pose substantial ambiguities regarding ownership of AI-generated content. When work is created by AI using an algorithm or datasets provided by a human, who should own the rights over the work that is created? Such problems underscore the need for an innovative approach to resolving intellectual property issues in the context of AI-generated material.
5. Conclusion
The article looks into the issues of sorting out IP laws for works that are made using artificial intelligence. For years, creative works were recognized as owned by people, but AI is now able to produce them, so this law does not reflect reality. According to Indian law, computer-generated works are accepted, yet Section 2(d)(vi) of the Copyright Act[32] limits authors of these works to real people. Such cases are not limited to the United States, as they can happen anywhere. While USPTO emphasizes that AI can be considered an author, Zarya states that view is not supported by the law. Human rights are important in IP for the EU and the U.S., but WIPO believes AI-built work needs a system that is different from others. There are ownership questions when AI can work on its own, since this raise concerns over who should hold the rights—the developer, the user, or no one. The writer suggests setting up IP rights for AI content that protect the human side of the project more than the computer element. There is a need for governments across the globe to agree and introduce new policies to tackle this developing issue.
[1] Azhan Saleem, ‘Legal Personhood of Artificial Intelligence and its Implications on Copyright Law’ (NLIU Journal of IP Law, 16 December 2024) <https://csipr.nliu.ac.in/miscellaneous/legal-personhood-of-artificial-intelligence-and-its-implications-on-copyright-law/> accessed 10 June 2025.
[2] JAG Griffith, ‘Review of Jurisprudence, by J Salmond and GL Williams’ (1947) 10(4) The Modern Law Review <http://www.jstor.org/stable/1089881> accessed 10 June 2025.
[3] The Copyright Act 1957, s 2(d)(vi).
[4] Rupendra Kashyap v Jiwan Publishing House [1996] (38) DRJ 81.
[5] Nikhil Mishra and Digvijay Singh, ‘AI-Generated Work and its Implications on Copyright Law in India’ (2025) 30(1) JIPR <https://or.niscpr.res.in/index.php/JIPR/article/download/5862/3865/66349> accessed 10 June 2025.
[6] Walter v Lane [1900] AC 539.
[7]Eastern Book Company & Ors v D.B. Modak & Anr [2008] (1) SCC 1.
[8] Feist Publications, Inc. v Rural Tel. Serv. Co., [1991] 499 U.S. 340, 345.
[9] R. Mehta and S. Soni, ‘Intellectual Property Rights, Ethical and Legal considerations in AI driven Innovations: Bridging Legal Gaps’ (2024) 11(2) International Journal of Research and Analytical Reviews <https://www.ijrar.org/papers/IJRAR1DOP007.pdf> accessed 10 June 2025.
[10] Stephen L. Thaler v. Comptroller General of Patents, Designs and Trademarks, [2021] UKSC/0201.
[11] Office of the Controller General of Patents, Designs and Trade Marks, ‘Draft Guidelines for Examination of Computer Related Inventions (CRIs)’ <https://www.ipindia.gov.in/writereaddata/Portal/Images/pdf/Draft_CRI_Guidelines_Publication_March2025.pdf> accessed 10 June 2025.
[12] Harshal Chhabra and Arihant Sethia, ‘The Impact of Artificial Intelligence on Intellectual Property Rights: A Case for Reform in Indian Patent Law’ (IJLT, 27 May 2024) <https://www.ijlt.in/post/the-impact-of-artificial-intelligence-on-intellectual-property-rights-a-case-for-reform-in-indian-p> accessed 10 June 2025.
[13] Som Prakash Rekhi v Union of India & Anr [1981] (1) SCC 449.
[14] The Copyright Act 1957, s 2(d)(vi).
[15] Harshal Chhabra and Arihant Sethia, ‘The Impact of Artificial Intelligence on Intellectual Property Rights: A Case for Reform in Indian Patent Law’ (IJLT, 28 November 2024) <https://www.ijlt.in/post/the-impact-of-artificial-intelligence-on-intellectual-property-rights-a-case-for-reform-in-indian-p> accessed 10 June 2025.
[16] United States Copyright Office, Zarya of the Dawn (Registration # VAu001480196, 21 February 2023) <https://www.copyright.gov/docs/zarya-of-the-dawn.pdf> accessed 10 June 2025.
[17] Tushar Dixit, ‘Intellectual Property Rights (IPR) in the age of Artificial Intelligence (AI) focusing on Indian Law’s’ (2025) 11(2) IJRA <https://www.allresearchjournal.com/archives/2025/vol11issue2/PartA/11-2-15-946.pdf> accessed 10 June 2025.
[18] Th. Tzimas, ‘Evolution of Copyright in the Era of Artificial Intelligence: Analysis of Conflicts of Law and Judicial Precedents’ (2025) 3(1) JDTL <https://www.lawjournal.digital/jour/article/view/508> accessed 10 June 2025.
[19] Thaler v Vidal [2022] No 21-2347 (Fed Cir).
[20] Berne Convention for the Protection of Literary and Artistic Works (signed 9 September 1886).
[21] Copyright Act of 1976, 17 U.S.C. §§ 101–1332 (2012).
[22] Nicholas, ‘Copyright Office Says AI-Generated Works Based on Text Prompts Are Not Protected’ (2025) 15 (161) NLR <https://natlawreview.com/article/copyright-office-says-ai-generated-works-based-text-prompts-are-not-protected> accessed 10 June 2025.
[23] National Artificial Intelligence Initiative Act of 2020, H.R. 6216, 116th Cong. (2020).
[24] Future of Artificial Intelligence Innovation Act of 2024, S. 4178, 118th Cong. (2024).
[25] World Intellectual Property Organization, WIPO Technology Trends 2019: Artificial Intelligence (WIPO 2019) <https://www.wipo.int/edocs/pubdocs/en/wipo_pub_1055.pdf> accessed 10 June 2025.
[26] Francis Gurry, ‘Artificial Intelligence and Intellectual Property: An Interview with Francis Gurry’ (2018) 5 WIPO Magazine <https://www.wipo.int/wipo_magazine/en/2018/05/article_0001.html> accessed 10 June 2025.
[27] Bureau Marocaine du Droit d’Auteur, ‘WIPO Dialogue on Intellectual Property and Artificial Intelligence: Second Session – Comments from Morocco’ (WIPO, 2019) <https://www.wipo.int/export/sites/www/about-ip/en/artificial_intelligence/call_for_comments/pdf/ms_morocco_bmda.pdf> accessed 10 June 2025
[28] Ravijaa Mehta and Shagun Soni, ‘Intellectual Property Rights, Ethical and Legal Considerations in AI Driven Innovations: Bridging Legal Gaps’ (2024) 11(2) IJRAR <https://ijrar.org/papers/IJRAR1DOP007.pdf> accessed 10 June 2025.
[29] European Commission, ‘AI Code of Practice’ (Digital Strategy, 2025) <https://digital-strategy.ec.europa.eu/en/policies/ai-code-practice> accessed 10 June 2025.
[30] Firas Massadeh, ‘The Legal Protection of Artificial Intelligence-Generated Work: The Argument for Sui Generis over Copyright’ (2024) 6(1) CLGR <https://www.researchgate.net/publication/378122622_The_legal_protection_of_artificial_intelligence-generated_work_The_argument_for_sui_generis_over_copyright> accessed 10 June 2025.
[31] Act on the Promotion of Research, Development and Utilization of Artificial Intelligence-Related Technologies (Japan AI Bill) (submitted 28 February 2025, passed House of Representatives 24 April 2025, pending enactment) (Japan).
[32] The Copyright Act 1957, s 2(d)(vi).
Authored by: Ms. Sanvi Sanvi Pipada and Mr. Vasu Srivastava
Ms. Sanvi Pipada and Mr. Vasu Srivastava are Third-Year Law Students at Dharmashastra National Law University, Jabalpur.
Leave a Reply