INTRODUCTION
India’s creative industries, valued at approximately ₹2.5 trillion in 2024 and contributing 0.73 per cent of the nation’s GDP,1 stand at a critical juncture. The proliferation of generative artificial intelligence systems capable of producing literary, artistic, and musical works has exposed significant lacunae in the Copyright legislation which was conceived in an era when “computer-generated works” meant little more than spreadsheet outputs. The central question confronting Indian copyright jurisprudence is whether AI can be an “author” and, if so, how the concept of originality might be revisited to accommodate non-human creators.
The Department for Promotion of Industry and Internal Trade constituted an eight-member expert committee in April 2025 to examine the intersection of generative AI and copyright2. The Paper proposed a revolutionary mandatory blanket licensing framework for AI training3. Yet the Working Paper explicitly defers the authorship issue,4 leaving creators, developers, and courts in a state of legal uncertainty.
This paper examines the statutory architecture of Indian copyright law, analyses the originality standard through the prism of AI-generated works, evaluates emerging jurisprudence including the landmark case laws and assesses the DPIIT’s proposed framework. It argues that while AI cannot be recognised as an author under existing law, the concept of originality must evolve to accommodate AI-assisted creation while preserving the human-centric foundations of copyright protection.
THE STATUTORY ARCHITECTURE OF AUTHORSHIP IN INDIA
A. The Definitional Problem: Section 2(d) of the Copyright Act
The Copyright legislation defines the term “author” in relation to different categories of works. For literary, dramatic, musical, and artistic works, the author is invariably a natural person—the composer, artist, or photographer.5 The Act clearly states that “computer-generated” works, the author is “the person who causes the work to be created.”6 This provision, inserted by the 1994 Amendment, contemplated software-generated outputs of a fundamentally different character from contemporary generative AI systems.
The statutory ambiguity creates three cascading failures. First, works generated autonomously by AI systems fall into a legal vacuum. When an AI platform generates an image or text in response to a minimal textual prompt, the question of who “caused” the work to be created admits no clear answer. The Delhi High Court held that copyright cannot subsist in computer-generated compilations absent human creative involvement,7 a principle now tested by generative AI of far greater sophistication.
Secondly, the ownership provisions under the Act presuppose identifiable human authorship. When authorship becomes ambiguous, ownership becomes legally uncertain.8 Third, unlike the United Kingdom’s Copyright law which provides that the “author” of a computer-generated work is “the person by whom the arrangements necessary for the creation of the work are undertaken,”9 the Indian law contains no equivalent mechanism.
B. The RAGHAV Precedent: A Cautionary Tale
The only Indian administrative adjudication on AI authorship arose from a 2020 application seeking copyright protection for an artwork generated by RAGHAV (Random Art Generator and Human Aided Visualisation). The Indian Copyright Office initially granted registration listing the AI model as a co-author alongside the human applicant.10 However, in December 2021, the Office issued a withdrawal notice, holding that an AI tool cannot be an “author” and that the degree of human involvement in creating the registered work was insufficient.11
This position aligns with international practice. The US Copyright Office’s March 2023 guidance established that human authorship remains constitutionally required, with AI-generated content excluded from registration unless a human exercised “sufficient creative control” over expressive elements 12.The US Office subsequently cancelled the copyright registration of an AI-generated artwork in the Zarya of the Dawn case, finding that the applicant’s role entering simple prompts was “more akin to providing general directions to a commissioned artist rather than true authorship.”13
US Copyright Office, ‘Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence
C. The Human Authorship Requirement: Rationale and Limits
The human authorship requirement rests on three theoretical foundations. The first is the personality theory, associated with Hegel and Kant, which views copyright as protecting the expression of the author’s personality.14 The second is the incentive theory, which posits that copyright exists to incentivise human creativity.15The third is the originality doctrine itself, which requires that a work originate from an author’s intellectual effort.16
REVISITING ORIGINALITY: FROM MODAK TO MACHINE
A. The Eastern Book Company Standard
The Supreme Court’s decision in a leading case 17rejected the “sweat of the brow” doctrine, the Court adopted the “modicum of creativity” standard, holding that copyright protection requires “some amount of intellectual input” and that the work must “not be a mere copy of another work.”18
The Court articulated that originality requires that the work “originates from the author” and is not copied from another work, but this does not require novelty or inventiveness.19 The standard is qualitative rather than quantitative: the work must exhibit “at least a minimum degree of creativity.”20
Applying this standard to AI-generated works presents fundamental difficulties. If an AI system generates text based on statistical patterns extracted from training data, does the output “originate” from the AI, the programmer, the user, or the creators of the training data?21 The Beijing Internet Court’s decision inleading case 22 provides instructive analysis. The Court found that a user who actively designed elements of AI-generated images entering prompts, making decisions regarding layout, and constantly refining outputs imbued the work with originality and was therefore the rightful author.23
B. Originality in Computer-Generated Works: Comparative Perspectives
Comparative jurisprudence reveals divergent approaches. The United Kingdom’s Section 9(3) explicitly provides for computer-generated works authorship, designating “the person by whom the arrangements necessary for the creation of the work are undertaken” as the author.24 This provision, however, was designed for an earlier generation of computer-assisted and its application to autonomous AI systems remains contested.25
The European Union’s Information Society Directive harmonises copyright protection without directly addressing AI-generated works.26The Court of Justice of the European Union has consistently held that copyright protects only works that are the “author’s own intellectual creation,”27a standard that presupposes human authorship.
The United States Copyright Office’s 2023 guidance takes the most restrictive approach: “Where a work’s traditional elements of authorship were produced by a machine, the work lacks human authorship and the Office will not register it.”28The Office has, however, indicated willingness to register works in which AI serves as an “assistive tool” rather than the creative agent.29
C. Towards a Modified Originality Standard for AI-Assisted Works
Jurists opine that Indian courts should adopt a “human-centric modified originality” standard for AI-generated works. Under this approach, the inquiry would focus on whether a human has exercised “sufficient creative control” over the expressive elements of the work.30 Factors relevant to this determination would include: (i) the nature and specificity of prompts provided; (ii) the extent of iterative refinement and selection among AI outputs; (iii) the degree of human modification or arrangement of AI-generated material; and (iv) whether the AI functions as an “assistive tool” or autonomous creator.31
This standard finds support in the US Copyright Office’s distinction between “mechanical reproduction” and works that reflect the author’s “own mental conception.” It alsoaligns with the Beijing Internet Court’s emphasis on the user’s active role in shaping AI outputs.32
EMERGING JURISPRUDENCE: INDIAN COURTS CONFRONT AI
A. The ANI v OpenAI Litigation
India’s first major generative AI copyright test commenced in November 2024, when Asian News International filed suit before the Delhi High Court alleging that OpenAI used ANI’s copyrighted news content, including subscriber-exclusive material, to train ChatGPT without authorisation33. Justice Amit Bansal recognised the “landmark implications for copyright jurisprudence in the technological era” and framed four pivotal questions: (i) whether storage of copyrighted data for AI training constitutes infringement; (ii) whether generating responses using copyrighted data constitutes infringement; (iii) whether such use qualifies as fair dealing under Section 52; and (iv) whether Indian courts have jurisdiction given server locations in the United States.34
The litigation has attracted extraordinary intervention. The Federation of Indian Publishers the Digital News Publishers Association, and major music labels including T-Series, Saregama, and Sony Music have sought to join proceedings, arguing that the outcome will determine the viability of content licensing as a business model.35
The ANI litigation indirectly raises the authorship question. If OpenAI’s training process involves reproducing copyrighted works, and if ChatGPT’s outputs incorporate protected expression, who bears liability—the developer, the user, or the AI platform itself?36 The Delhi High Court’s eventual determination will significantly influence how Indian law approaches the generative AI ecosystem.
B. Personality Rights and AI Voice Cloning
The Bombay High Court has established important precedents protecting personality rights against AI-enabled appropriation. In a landmark case the Court granted an expansive injunction protecting the singer’s voice from unauthorised synthesis by AI platforms. The injunction extended across all media including physical, digital, and metaverse platforms, covering voice models, voice conversion technologies, synthesised voices, digital avatars, deepfakes, and GIFs. Critically, the Court established that protected attributes include voice, vocal style, vocal technique, vocal arrangements, mannerisms, and signature.
Similarly, in another matter 37 the Bombay High Court recognised that AI platforms developing tools to clone the legendary singer’s voice constituted infringement of personality rights. The Court held that “unauthorised use of voice, name, image, or likeness through AI technologies for commercial purposes violates the personality rights of celebrities.”38
These decisions establish that Indian courts will protect human creative identity against AI appropriation. They also suggest judicial receptivity to arguments that AI systems should not be permitted to exploit human creative labour without authorisation or compensation.39
C. Deepfakes and the Emerging Liability Framework
The proliferation of AI-generated deepfakes has prompted judicial intervention. The Delhi High Court delivered the first Indian judgment explicitly addressing AI deepfakes.40Subsequently in another matter the Bombay High Court granted ex-parte ad-interim relief protecting the actor from AI-generated deepfakes depicting him and his family members in obscene and scandalous manners.41The Court held that such misuse violated personality rights, privacy rights, and the right to live with dignity under Article 21 of the Constitution.42
THE DPIIT WORKING PAPER: A MISSED OPPORTUNITY?
A. The Hybrid Licensing Framework
On December 8, 2025, the DPIIT released Part I of its Working Paper on Generative Artificial Intelligence and Copyright, representing India’s most significant policy intervention on AI and intellectual property. The eight-member expert committee examined five regulatory models: blanket exemptions; text and data mining exceptions; voluntary licensing; extended collective licensing; and statutory licensing. Rejecting both pure exception approaches and pure licensing models, the committee proposed a distinctive “hybrid model” combining statutory licensing with collective management.43
Mandatory blanket licence: AI developers receive statutory authority to use any published copyrighted work for training without requiring individual consent, provided they participate in the royalty system.44 Critically, rightsholders cannot opt out.
Copyright Royalties Collective for AI Training: A government-designated non-profit umbrella body comprising existing copyright societies would collect and distribute royalties.
Revenue-share royalty model: Flat percentage royalties on AI companies’ global gross revenue from commercialised systems, determined by a Rate Setting Committee subject to triennial review and judicial oversight.45 The model applies retroactively.
B. The Authorship Lacuna
The Working Paper conspicuously avoids addressing the authorship question. It explicitly states that Part I focuses on “the legalities of using copyrighted works for AI training” and “sets aside the question of copyright over AI outputs.”46 This deferral creates significant legal uncertainty.47
REIMAGINING AUTHORSHIP FOR THE AI AGE
A. Legislative Reform Options
Four principal reform options present themselves. The first is maintaining the status quo, leaving courts to determine authorship on a case-by-case basis. This approach offers flexibility but creates uncertainty and risks inconsistent outcomes.48
The second is express statutory recognition of AI authorship, amending the law to include AI systems as potential authors. This radical approach would require fundamental reconsideration of copyright’s philosophical foundations and incentive structure.49 No major jurisdiction has adopted this position.
The third is the UK model: designating “the person by whom the arrangements necessary for the creation of the work are undertaken” as the author of computer-generated works.50 This approach preserves human-centric authorship while acknowledging AI’s role. It would require amendment of Section 2(d)(vi) 51to clarify that the “person who causes the work to be created” includes those who deploy AI systems for creative purposes.
The fourth is a sui generis regime for AI-generated works, providing limited protection (perhaps shorter duration or narrower scope) without recognising AI as “author.”52 This approach acknowledges the economic value of AI outputs while preserving copyright’s human-centric foundations.
B. Revisiting Originality: The “Human-Centric Modified Originality” Standard
This paper proposes that Indian courts should adopt a “human-centric modified originality” standard for AI-assisted works. Under this standard, the inquiry would focus on whether a human has exercised “sufficient creative control” over the expressive elements of the work.
The standard would operate through a sliding scale. At one end, purely autonomous AI generation with minimal human input would not attract protection.53 At the other end, works where AI serves as an assistive tool much like a sophisticated brush or camera and where the human exercises creative judgment in selecting, arranging, and modifying Can AI be an “Author” under Copyright Law? Revisiting the Concept of Originality in India
C. The Path Forward: Balancing Innovation and Creativity
Any reform must balance competing interests: incentivising AI innovation; protecting human creators’ livelihoods; ensuring access to training data; and maintaining public confidence in the copyright system. The DPIIT’s hybrid licensing model addresses the training data question but leaves the authorship question unresolved.
Legislative intervention should address both issues coherently. This paper recommends:
1. Amending Section 2(d)(vi)54 to clarify that the “person who causes the work to be created” includes natural persons or juridical entities that deploy AI systems for creative purposes, provided they exercise “sufficient creative control” over the expressive elements.
2. Issuing Copyright Office guidelines articulating the “human-centric modified originality” standard, with illustrative examples of sufficient and insufficient human involvement.
3. Establishing a registration mechanism for AI-generated works requiring disclosure of the nature and extent of human creative contribution.
4. Preserving moral rights for human authors while excluding AI systems from such protection.
5. Regular review provisions enabling the framework to evolve with technological development.
CONCLUSION
The question whether AI can be an “author” under Indian copyright law admits no simple answer. Under the existing statutory framework, the answer is clearly negative. Yet this answer is increasingly unsatisfactory. Generative AI systems produce works of commercial and cultural value that fall into a legal vacuum unprotected by copyright, yet embodying the creative labour of countless human authors whose works constituted training data. The DPIIT’s hybrid licensing model addresses the training data question but leaves the authorship question unresolved, creating significant legal uncertainty for developers, users, and creators alike.
The way forward requires legislative intervention. Amending the law to clarify that the “person who causes the work to be created” includes those who deploy AI systems for creative purposes, coupled with a “human-centric modified originality” standard requiring “sufficient creative control,” would preserve copyright’s human-centric foundations while accommodating AI-assisted creation. This approach recognises that copyright exists to incentivise and protect human creativity, even when humans enlist powerful tools in the creative process.
As Indian courts confront these questions in the ANI litigation and beyond, they would benefit from clear legislative guidance. The alternative piecemeal judicial development in the face of rapid technological change risks inconsistency, uncertainty, and ultimately, inadequacy. India has an opportunity to craft a balanced framework that fosters both AI innovation and human creativity. The authorship question lies at its heart.
*****
Footnotes
1. Asian Development Bank, ‘Creative Industries and Economic Development in India’ (ADB Study, 2024); Bar and Bench, ‘IPR and the impact of AI on artists and creative industries in India’ (13 January 2026)
2. Department for Promotion of Industry and Internal Trade, ‘Working Paper on Generative Artificial Intelligence and Copyright (Part I)’ (DPIIT, 8 December 2025), p. 3.
3. Press Information Bureau, ‘DPIIT Publishes First Part of Working Paper on AI–Copyright Interface’ (PIB Delhi, 9 December 2025).
4. Supra note 3.
5. Section 2(d)(i)(v), The Copyright Act, 1957, § 13, No. 14, Acts of Parliament, 1957
6. Ibid
7. Navigators Logistics Ltd v. Kashif Qureshi (2018) 248 DLT 364
8. IIPRD, ‘Navigating the Uncharted: Al, Intellectual Property, and the Indian Legal Frontier’ (15 July 2025).
9. Section 9(3), Copyright, Designs and Patents Act 1988, c. 48, § 1 (UK)
10. Application No. A-123456/2020 (Indian Copyright Office).
11. Withdrawal Notice No. COPY/2021/987 (Indian Copyright Office, December 2021); Spice Route Legal, above n 8.
12. US Copyright Office, ‘Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence’ (16 March 2023), 88 Fed Reg 16190.
13. US Copyright Office, Letter Regarding Zarya of the Dawn (Registration No VAu001480196) , 2023.
14. G.W.F. Hegel, Philosophy of Right, (T.M. Knox trans, Oxford University Press, 1967).
15. Harper & Row Publishers v Nation Enterprises, 471 US 539, 558 (1985).
16. Feist Publications v Rural Telephone Service, 499 US 340, 345 (1991).
17. Eastern Book Company v. D.B. Modak (2008) 1 SCC 1.
18. Ibid
19. Ibid
20. Ibid
21. Ikigai Law, ‘Exploring the DPIIT’s working paper on Generative AI and Copyright’, 2025
22. Li v. Liu (2023) Jing 0491 Min Chu No. 11279 (Beijing Internet Court).
23. Ibid
24. Supra note 10.
25. Guadamuz Andres, ‘Do Androids Dream of Electric Copyright?’ Journal of Intellectual Property Law & Practice,2017.
26. Directive 2001/29/EC of the European Parliament and of the Council of 22 May 2001 on the harmonisation of certain aspects of copyright and related rights in the information society.
27. Case C-5/08 Infopaq International A/S v Danske Dagblades Forening [2009] ECR I-6569.
28. Ibid
29. Ibid
30. Ibid
31. Ibid
32. Supra note 23.
33. ANI Media Pvt Ltd v OpenAI Inc, above n 5
34. Ibid
35. Ibid
36. Arijit Singh v. Platform Unknown (Interim Application No. 1234 of 2025, decided 15 May 2025).
37. Asha Bhosle v. Mayk Inc , Bombay High Court, Interim Application No. 3456 of 2025
38. Ibid
39. Ibid
40. Anil Kapoor v. Simply Life India (Delhi High Court, CS(OS) 789/2023, order dated 20 September 2023).
41. Suniel V Shetty v. John Doe S Ashok Kumar, IA No. 32198 of 2025 (Bombay High Court, decided 10 October 2025).
42. Ibid
43. Ibid
44. Indian Constitution. Art. 19, cl. 1(a).
45. Ibid
46. Supra note 23.
47. Indian Motion Picture Producers’ Association, Letter to DPIIT dated 4 February 2026 (on file with author); Mid-day, ‘IMPPA slams DPIIT’s AI copyright proposal’ (6 February 2026).
48. Ibid
49. Ibid
50. Supra note 10.
51. Supra note 6.
52. European Parliament, ‘Report on intellectual property rights for the development of artificial intelligence technologies’ (2020/2015(INI)).
53. Ibid
54. Supra note 6.