Abstract
Generative Artificial Intelligence (AI) has revolutionized the production, reproduction and dissemination of creative works, as well as posing intricate challenges for copyright law. A major type of synthetic media, deepfakes also pose further issues regarding copyright violations, personality rights, privacy concerns, and intermediary liability. While the Copyright Act, 1957 in India contains a statutory framework to address computer-generated works, it does not provide a clear answer on authorship and ownership issues for AI-generated works. In recent developments, in 2026, the Copyright Office took a stance on the issue of AI authorship and the Information Technology Rules on Synthetically Generated Information have also been updated. The article explores the legal landscape surrounding AI-generated works and deepfakes in India, outlines existing challenges, and reviews the importance of establishing a fair balance to accommodate technological advancements while safeguarding creators and individuals.
1. Introduction
AI has revolutionized the way creative content is created, reproduced and distributed. Today, generative AI can create images, music, videos, literary works and even very realistic digital avatars. One of the most controversial applications of this technology is the creation of synthetically generated or manipulated voice and video that makes a person look, sound, or act like as if they had done or said something that they actually did not. As AI-generated content becomes more prevalent, there is a complex challenge for copyright law: who has the copyright to a work created with AI’s help and what is the situation if AI is used to reproduce or manipulate a copyrighted work? Where a deepfake contains the face, voice and performance or other recognisable traits of a person, the situation is made even more complex. There is currently no single comprehensive legislation to regulate AI-generated content and deepfakes in India as of 2026. Instead, the legal framework is spread out through the Copyright Act, 1957, the Information Technology Act, 2000 and judicially created personality and privacy rights, which were further updated in the Information Technology Rules, 2021 (as amended in 2026).
2. Understanding AI and Deepfake
AI-generated content is content that is created entirely or partially by AI systems. Synthetic media is a subset of a deepfake where AI is employed to generate or alter voice, images or video to make them seem true. For instance, an AI system can create a video of a famous actor making a declaration that the actor did not make. Or the system can generate a piece of art that is very similar to an artist’s current style. There are two different questions of law that arise in these circumstances. Firstly, if the content created by the AI itself is eligible for copyright. Second, whether that creation or dissemination is a copyright, privacy, publicity or other right violation of another person.
3. Copyright Protection Under Indian Law
The point of departure is the Copyright Act, 1957. Copyright in original literary, dramatic, musical and artistic works, cinematograph films and sound recordings is granted in Section 13. Section 14 provides copyright owners with exclusive rights such as: reproduction, communication to the public, adaptation, and others. An important provision in the Act is one which specifically refers to computer-generated works. The author of a literary, dramatic, musical or artistic work, which is computer-generated, is the person who causes the creation of the work.
This is a very pertinent provision during the age of artificial intelligence. It will help to define the scope of Indian copyright law for computer-generated works, but will not explicitly provide solutions to all the queries arising with the current generative AI. For instance, does the user become the “person who causes the work to be created” when providing a detailed prompt to an AI system and that system independently creating an image? Or because human creativity is lacking, is it not possible for there to be copyright? The issues raised in these questions continue to be the main concerns in the field of Indian AI copyright law.
4. AI Generated Works and the Question of Authorship
A significant event took place in 2026, when the Indian Copyright Office received an application related to an AI-created artwork in connection with a DABUS system. The Copyright Office is said to have agreed that the work was original but denied that the AI system itself could be an “author” of a work under Section 2(d)(vi) because an AI system is not a natural or juristic person that can be recognised as an author of a work. This is very important. It implies that within the current system, as is, AI alone cannot be considered the legal author simply because it created the work. The thornier issue is who is the human being that the statute requires to be the author of the work?
Therefore, the stronger argument is for copyright to be based on identifiable human involvement, not the mere AI system’s output. The 2025 DPIIT Working Paper on Generative AI and Copyright too illustrates that the Government has acknowledged that there are still undecided issues regarding the training of AI tools and the copyrightability of AI-generated content.
5. Deepfakes and Copyright Infringement
Not everything is a deep fake is not an infringement on copyright. The legal analysis will rely on the type of creation of the deepfake and the way in which the material was employed. Let’s say a person takes a photo of a celebrity that is copyrighted and uses AI to manipulate the image into a video. The copyright owner has his exclusive rights under Section 14 of the Copyright Act involved if the copyrighted photograph or other protected material is reproduced or substantially incorporated without his authorisation. But there are limitations to what constitutes copyrightable aspects of an individual’s identity. The issues of privacy and personality/publicity rights are distinct when it comes to a person’s face, name, voice or persona.
This difference is very significant, copyright covers original creative expression, personality rights cover identifiable facets of an individual’s personality. Therefore, a deep fake of a celebrity can generate copyright, privacy and passing off claims as well as claims for personality rights at the same time, if at all, depending on the circumstances. In some circumstances the fair dealing provisions of the Copyright Act (Section 52) may also be relevant, such as criticism, review or reporting on current events.
6. The 2026 Regulatory Position on Deepfakes
In 2026, India’s regulatory landscape took a significant turn with the updates to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. The revised framework has new obligations on Synthetically Generated Information (SGI). The framework is dedicatedly designed for synthetic audio, visual and audio-visual content, such as deepfakes. The framework emphasizes a significant part of the role of intermediaries and platform duties, such as on the identification, labelling, traceability and removal of illegal synthetic content.
This is significant since it takes India’s approach beyond simply general laws on criminality and intermediaries to tackle the issue of deepfakes. The Government has explicitly acknowledged the dangers of synthetic media, such as impersonating, misinformation, privacy concerns and reputation damage. But these guidelines focus mostly on platform regulation and damaging synthetic material. They don’t establish a comprehensive statutory copyright regime for AI works.
7. Judicial Developments and Personality Rights
In recent years, Indian courts have increasingly taken up the challenge of the misuse of AI and digital technologies in the context of personality rights. In recent times, the Delhi High Court has shown its willingness to curb unauthorised use of an individual’s name, image, voice, likeness, persona and AI generated or deep fake content, as illustrated in the case Yuvraj Singh v. Ashok Kumar & Ors.[1], in which the Delhi High Court Bench of Jyoti Singh granted an ex parte and interim injunction protecting the personality and publicity rights of former Indian cricketer Yuvraj Singh on 29th July 2026.
In the 2026 order, the Court explicitly addressed the issues of deepfake content, manipulated content, and AI generated content in safeguarding personality and privacy rights. What’s important about these decisions is that the victim of a deepfake does not necessarily need to turn to copyright law, as is often suggested. Judicial relief may be granted on an independent basis by the personality rights and privacy principles. This is especially relevant when the subject of the original photo or video is not the copyright holder of the image or video.
Conclusion
2026 positions of the legal sphere in India could best be described as developing and not defined. Some principles are put in place under the Copyright Act, through its Section 2(d)(vi) but this was not constructed to deal with generative AI. The Copyright Office’s 2026 position clearly defines that AI per se can’t be the author, human contributions continue to be the central component of copyright authorship. On the other hand, the changes introduced to the Indian IT rules have provided an even clearer legal basis for dealing with synthetically produced information and deepfakes, along with recent developments on personality and privacy rights to complement such protection.
Indian law thus currently needs to face the dilemma between protecting human authorship, preventing the abuse of an individual’s identity and providing necessary support for the advancement of technological innovation. With increasing pervasiveness of AI within the creativity domain, it’s essential that law moves away from only asking “Who created the work?” to a deeper question; “What role did humans contribute and in what proportion did machines contribute to the creation process?”.
Author: Akshi Seem, Associate Partner and Priyanshu, Intern
- CS(COMM) 796/2026.
