2025 COPYRIGHT REFORMS: DUAL IMPACT OF DIGITAL PAYMENTS AND AI POLICY ON INDIA’S CREATIVE ECONOMY

inbatv Law Students, Legal

INTRODUCTION
Generative AI has caused profound changes in how we create, consume and monetize creative content and thus has affected the Copyright Laws. The introduction of Rule 83A has caused a paradigm shift and the proposed changes to AI training and digital licensing has resulted in new copyright regulation environment. This article will review how the evolution of digital payment changes will impact AI copyright law in the Indian Creative Industry.

DIGITAL PAYMENTS AND RULE 83A- PROCEDURAL REFORM IN INDIAN COPYRIGHT RULES

Prior to the amendment of 2025, Rule 83’s focus was inclined towards infringement remedies rather than efficient remuneration mechanisms. The copyright licensing and royalty payments in India were traditionally opaque, fragmented and dependent on manual systems and ample adversities faced by the creators entail delayed payments, lack of transparency and dependence on intermediaries. Rule 83A which was introduced on June 4th, 2025 as an amendment to the Copyright Rules, 2013, emerges against this backdrop as an attempt to integrate digital payment systems into copyright licensing. As per Rule 83A, 

“the owner or licensor of a literary work, musical work, and sound recording shall establish and maintain an online payment mechanism for the collection of license fee payable by a licensee for communication to the public of such work. All payments of such license fee shall be processed exclusively through said online system, and no alternative method of payment shall be permitted or accepted for this purpose.”

This rule thus mandates copyright owners or licensors of literary works, musical works, and sound recordings to establish and maintain an online system for collecting licence fee from licensees but it limits the form of transaction to digital transactions exclusively. This amendment signals a paradigm shift in copyright governance from enforcement-centric mechanisms to payment-centric regulation. Unlike Rule 83, in which fee is payable to Copyright Office, Rule 83A specifically addresses the collection of licence fees by copyright owners or licensors.

This rule has impact majorly on three groups:  creators, licensees and intermediaries. For creators, it promises greater transparency and predictability in royalty payments. From the perspective of licensees or digital platforms, the rule offers clarity and standardization in licensing procedures. Simultaneously, the intermediaries face increased scrutiny and a diminished scope for discretionary practices, as digital records make transactions more easily auditable.

Despite its reformist intent, this rule is critiqued for placing the procedural obligations on licensors, ignoring licensees and intermediaries who play prominent role in payment flows. Moreover, the rule does not extend to other licensing domains such as adaptation, translation, or synchronization, thus limiting it applicability in the digital and AI-centric era.

AI AND COPYRIGHT POLICY DEBATE

The Copyright Act, 1957 was drafted long before generative AI existed. The core rights which are reproduction, adaptation, storage, communication to public are the exclusive rights of creators. AI training involves downloading and storing works, transforming them into datasets and using them to produce outputs and such acts prima facie infringe copyright laws unless there is a licence from the rightsholder, a fair dealing defence or a statutory exception as given under Section 52 of the Copyright Act, 1957. 

The Department for Promotion of Industry and Internal Trade’s (DPIIT) working paper proposes a hybrid compulsory licensing model leveraging the Government policy ‘One nation, One Licence, One Payment’ which aims to centralize rights to use all ‘lawfully accessed copyrighted content’ for AI training, payment of royalties through a single entity, and distribution of those royalties to creators. Under this model, the AI companies will obtain a blanket licence from a central body designated by Government and the companies need not negotiate individual licences with every creator which is even impractical as AI extracts data from millions of sources. A central body like Copyright Royalties Collective for AI Training (CRCAT) would collect royalties from AI developers, maintain a registry of works for AI training, and distribute royalties to both members and non-members. Moreover, the authors would receive statutory remuneration whenever their work is included in AI training.

In the ongoing litigation ANI Media Pvt. Ltd. v. OpenAI Inc., ANI, a leading news agency alleges that OpenAI uses its copyrighted news content to train ChatGPT, thereby amounting to infringement. OpenAI has invoked doctrines such as fair dealing and questioned territorial jurisdiction. The working paper of DPIIT explicitly rejects the statutory exemption similar to text and data mining (TDM) exception found in foreign jurisdictions such as GEMA v OpenAI and Robert Kneschke v. LAION , these regimes were not intended for generative systems that replace creative markets, but rather for analytical data mining. According to DPIIT, zero-price statutory exceptions run the risk of allowing large tech companies to profit from creative works without paying for them, which would destabilize India’s creative economy. As a result, India’s policy decision supports a licensing-based framework that harmonizes AI innovation with copyright law’s economic and constitutional goals.

LEGAL AMBIGUITIES: FAIR DEALING, TRAINING DATA & COPYRIGHT LAW

The very essence of AI training requires the mass reproduction and storage of copyrighted works, which are converted into machine-readable datasets used to generate outputs. This means that copying is not incidental to AI but structurally essential to how models are built. However, the core problem that arises is that Indian law is yet silent on AI training.

While AI developers contend that training qualifies as research under Section 52, Dr. Torrance and Dr. Tomlinson contradict the statement by contending that AI training is not merely reading as done by humans, but it entails systematically copying, storing and processing entire works to generate outputs that may substitute for original works. Concomitantly, the authors have introduced the concept of ‘fair training’, asserting that AI training is transformative, non-expressive, and non-consumptive because it extracts patterns rather than expressive content.

Nevertheless, this remains a contentious issue from a legal point of view. In Indian fair dealing provisions, it is evaluated on a case-by-case basis, based on the nature of the purpose, use, and effect of such use on the primary work. While there is certainly a transformative nature of AI learning processes, it is found that generative processes, especially, have resulted in producing works which compete with books, news reports, illustrations, or screenplay books, thereby creating a substitute. Thus, it is hard to justify the applicability of Section 52 to Big Data-driven AI learning and thus this uncertainty explains the logic behind licence-based model rather than statutory exceptions.

IMPACT ON INDIA’S CREATIVE INDUSTRIES

Bollywood & Music Industry: Licensing and Royalty Stakes

India’s film and music industry including T-series, Saregama, Sony Music and Indian Music Industries generate their revenue from licensing and royalties. However, they face serious challenges from AI systems replicating their music, lyrics and audio content, and several large industry players have tried to intervene in copyright litigation cases brought against AI-related businesses (such as OpenAI) on grounds of training models on songs, lyrics, films, and recordings without authorization, infringing on their exclusive right.

Publishing Sector: Books, Literature & Market Substitution Risks

Indian authors see generative AI as posing unique threats and the Federation of Indian Publishers (FIP) criticized AI training practices that use copyrighted published content without permission and they demand not only limit on unauthorized access but also compensation mechanisms for past use.

News Media: Copyright Protection & Legal Actions

News media has become one of the most affected sectors by the use of generative AI since it involves the frequent ingestion of copyrighted news articles within training data sets to be used for crafting responses using AI. The Digital News Publishers Association has asserted that this violates their copyright, jeopardizing the feasibility of their business, since these tools have the potential to supplant original reporting, as submitted in the case of ANI Media Pvt. Ltd. v OpenAI Inc. pending in the Delhi High Court.

India’s Copyright Act’s blanket license-royalty structure will soon transform AI training into an organized, compensated sector in order to ensure fair licensing and compensation to the creators.

WAY FORWARD AND CONCLUSION

India’s 2025 copyright reforms are an important shift towards payment-centric approach. It is not inclined towards inclusion of statutory exceptions, but protecting creators and their creations. The introduction of digital payments under Rule 83A aims to bring transparency to the AI training development process by using a blanket licensing model which needs a centralized body to keep detailed registries of how works are licensed. To incorporate this, there is need for clear legal definitions of key terms, such as AI training, dataset disclosures, and liability for AI-generated works. If implementation of this model proves to be a success, it could help the global economy integrate AI into its creative sectors while supporting the economic sustainability of Bollywood, book publishing, and newspaper publishing.

[1] The Copyright Act, 1957 (Act 14 of 1957) read with The Copyright (Amendment) Rules, 2025, r. 83A. [1] Draft Copyright (Amendment) Rules, 2025 (India): Mandatory Online Payments for Licence Fees — Analysis, Medianama (11 June 2025) https://www.medianama.com/2025/06/223-draft-copyright-amendment-rules-2025-digital-payments/ accessed 15 January 2026.

 Working Paper on Generative AI and Copyright, Department for Promotion of Industry and Internal Trade, Ministry of Commerce & Industry (Dec 2025) https://www.dpiit.gov.in/static/uploads/2025/12/ff266bbeed10c48e3479c941484f3525.pdf accessed 15 January 2026.
 Exploring the DPIIT’s Working Paper on Generative AI and Copyright, Mondaq (15 Dec 2025) https://www.mondaq.com/india/copyright/1718746/exploring-the-dpiits-working-paper-on-generative-ai-and-copyright accessed 15 January 2026.
 ANI Media Pvt. Ltd. v. OpenAI Inc. & Anr., CS(COMM) 1028/2024 (Delhi High Court, India) (ongoing).
 GEMA v. OpenAI, Regional Court of Munich I, Case No. 42 O 14139/24 (11 Nov 2025) (Germany).

 Robert Kneschke v. LAION e.V., Case No. 5 U 104/24 (Higher Regional Court of Hamburg, Germany, 10 Dec 2025).

Andrew W. Torrance & Bill Tomlinson, Training Is Everything: Artificial Intelligence, Copyright, and Fair Training (2023) arXiv:2305.03720 https://arxiv.org/pdf/2305.03720 accessed 15 January 2026.

 The Copyright Act, 1957, s 52(1)(a) (India).
 Dr. Andrew W. Torrance, Ph.D. Paul E. Wilson Distinguished Professor of Law at the University of Kansas Visiting Scholar at the MIT Sloan School of Management
 Dr. Bill Tomlinson, Ph.D. Professor of Informatics at the University of California, Irvine Adjunct Professor, Te Herenga Waka – Victoria University of Wellington
 Andrew W. Torrance and Bill Tomlinson, Training Is Everything: Artificial Intelligence, Copyright, and Fair Training (2023) arXiv:2305.03720 https://arxiv.org/pdf/2305.03720 accessed 15 January 2026.
 T-Series, Saregama, Sony Music look to join copyright suit against OpenAI, Business Standard (14 Feb 2025) https://www.business-standard.com/companies/news/bollywood-t-series-saregama-sony-music-copyright-lawsuit-openai-delhi-hc-125021400876_1.html accessed 15 January 2026.
 ANI Media Pvt. Ltd. v. OpenAI Inc. & Anr., CS(COMM) 1028/2024 (Delhi High Court, India) (ongoing).
 Usage of copyright content: Digital news publishers join legal battle against OpenAI, Indian Express (27 January 2025) https://www.indianexpress.com/article/india/openai-case-indian-news-websites-copyright-9802312/ accessed 15 January 2026.

By: Roopkiran Kaur Shetra

College: University Institute of Legal Studies, Panjab University, Chandigarh

Course: B.Com.L.L.B. (Hons.) [3rd Year]