Summary
The DPIIT consultation paper titled ‘One Nation One License One Payment’ examines the use of copyright-protected works for training generative AI systems in India. It outlines a proposed hybrid framework combining a mandatory blanket licence with a statutory right to remuneration administered through CRCAT, including revenue-linked and retrospective royalty obligations, and records supporting and dissenting stakeholder submissions. The post also notes the notified mode and timelines for sending comments to DPIIT.
Background
The Department for Promotion of Industry and Internal Trade (DPIIT) has released Part I of a consultation paper on Generative Artificial Intelligence and Copyright, titled One Nation One License One Payment. The document addresses intellectual property concerns arising from the utilisation of protected creative works as training inputs for generative AI systems.
A committee constituted by DPIIT on 28 April 2025 prepared this first part of the consultation. The scope is limited to the deployment of copyrighted material for AI training purposes. Questions relating to copyrightability, authorship, moral rights, and liability connected with outputs generated by AI fall outside this phase and are proposed to be addressed in Part 2.
Copyright Law and AI Training
The paper states that the Copyright Act, 1957 does not contain an express provision addressing text and data mining or AI training activities. In the absence of such a provision, questions arise on whether these activities fall within the fair dealing framework under Section 52 of the Act.
Reference is made to ongoing judicial proceedings, including the matter of ANI Media Pvt. Ltd. v. Open AI Inc. before the Delhi High Court, where these questions are under consideration.
Policy Concerns Identified
According to the committee, generative AI systems are trained on large datasets that include works created by humans and protected by copyright. The deployment of such material for training purposes has given rise to concerns relating to authorisation, compensation, and the ability of rights holders to identify and manage how their works are being used.
The paper places these issues within India’s broader artificial intelligence policy landscape and refers to the country’s position as both a producer of creative works and a market for AI technologies.
Stakeholder Consultations
The committee consulted stakeholders from both the technology and content industries. According to the paper, representatives from the tech and AI industry advocated for a blanket Text and Data Mining (TDM) exception enabling training of generative AI on protected works. A group of stakeholders expressed support for a TDM exception with an opt out right for copyright holders. In a separate consultation, representatives of the content industry advocated for a voluntary licensing model.
Text and Data Mining and Other Regulatory Options
The committee examined regulatory approaches including voluntary licensing, statutory licensing, collective licensing mechanisms, blanket text and data mining exceptions, and exceptions accompanied by opt out or rights reservation mechanisms.
The paper states that voluntary licensing presents challenges regarding scalability and transaction costs. Broad exceptions for commercial purposes are described as raising concerns about protection and remuneration for creators. Opt out based approaches are described as dependent on effective transparency and rights reservation systems.
International Approaches
The paper refers to legal and policy frameworks across jurisdictions including the United States, Japan, the United Kingdom, the European Union, and Singapore. These regions have adopted different approaches to address text and data mining in the context of AI development, including statutory exceptions, interpretative doctrines, and licensing mechanisms.
The committee’s review indicates that no single approach has addressed all issues relating to scale, legal certainty, remuneration, and innovation in the context of commercial generative AI systems.
Hybrid Licensing Framework
The committee has proposed a hybrid licensing model under which AI developers would receive a mandatory blanket licence to utilise lawfully accessed protected works for training generative AI systems. The framework also provides a statutory right to remuneration for rights holders. Under this model, rights holders would not have the option to withhold their works from use in AI training.
Collecting Entity
The proposed structure provides for a centralised non profit entity called the Copyright Royalties Collective for AI Training (CRCAT). This entity would be formed by associations of rights holders and designated by the Central Government under statute. CRCAT would have Copyright Societies and Collective Management Organisations as its members, with one member for each class of works.
Royalty Rates and Distribution
The paper states that royalty rates would be determined by a government appointed Rate Setting Committee and would be based on a percentage of global revenue earned from commercialisation of AI systems trained on copyrighted content. The rates set would be subject to judicial review. The obligation to pay royalties would apply retroactively to AI developers who have trained their systems on protected content and are earning revenues from such systems.
The collecting entity would distribute royalties to both members and non members of Copyright Societies and CMOs who register their works for the purpose of receiving royalties related to AI training.
Stated Objectives
According to the paper, the framework aims to achieve the following: availability of all lawfully accessed copyrighted content for AI training without the need for individual negotiations; reduced transaction costs and compliance burden for AI developers; fair compensation to copyright holders; judicial review over royalty rates; simplified payment processes; mitigated risk of AI bias and hallucinations; and a level playing field for all stakeholders including startups.
Ministry of Electronics and IT Submission
The Ministry of Electronics and Information Technology (MEITY), in its submission to the committee, expressed support for the hybrid model. According to its submission reproduced in the paper, MEITY stated that the model has the potential to meet objectives across the domains of technological innovation and creative labour.
Dissenting View
Through its submissions dated 17 August 2025, Nasscom expressed dissent regarding the hybrid approach. According to the paper, Nasscom recommended a Text and Data Mining exception for both commercial and non commercial purposes where access is lawful and a good faith knowledge safeguard is met, solely for the training and input processing stage of machine learning. Nasscom proposed that rights holders should be able to reserve their works from TDM through machine readable opt out for publicly accessible content online, and through contract or licence terms for content that is not publicly accessible.
Submission Details
Comments and feedback on Part I may be submitted to DPIIT at ipr7-dipp@gov.in within 30 days of the notification dated 8 December 2025. The deadline for submissions is on or before 7 January 2026.
Disclaimer:
Parts of this blog post were generated with the assistance of an AI application. The views expressed are personal.