Generative AI and Copyright Issues

• 17 Jul 2025

For Prelims: Artificial intelligence, Copyright Infringement, ChatGPT, DU Photocopy Case (Oxford v. Rameshwari Photocopy Services, 2016) case 

For Mains: Copyright Infringement and Use of AI, Fair Use and Transformative Use in the Context of AI-Generated Works 

Source: IE 

Why in News?

The rapid growth of generative artificial intelligence (AI) like ChatGPT and Gemini has triggered concerns over the use of copyrighted content, raising key debates around intellectual property rights, authorship, deep fakes, and ethical AI governance.  

• These developments challenge conventional legal and ethical frameworks and require urgent attention. 

 

What is Artificial Intelligence (AI)? 

• Artificial Intelligence (AI), first coined by John McCarthy in 1956 broadly refers to the ability of machines to perform tasks that typically require human intelligence, such as learning, reasoning, and problem-solving. 
• Generative AI represents a major advancement in AI, enabling machines to create original content—such as text, images, music, code, or videos—by learning from large datasets like books, websites, and digital art.  
• Eg: ChatGPT, Gemini, and Claude for natural language generation; DALL·E and Midjourney for image creation; and AIVA and Amper Music for music composition. 

 

What are the Key Copyright Challenges Related to AI-generated Content? 

• AI Mimics Copyrighted Material: AI models require vast datasets, often comprising copyrighted texts, images, and music, to train and optimize performance, raising concerns of unauthorized copying when AI mimics or replicates creative elements of original works. 
  • Although the AI may not directly reproduce works, its outputs can closely resemble protected content, leading to potential copyright infringement. 
• Fair Use & Transformative Use: The doctrine of fair use allows limited use of copyrighted material without permission, especially for research, education, or commentary. In the US, it is based on 4 factors: purpose, nature, amount used, and market impact.  
  • Tech firms claim AI training is transformative use, adding new expression, meaning or utility and qualifying as “non-expressive use”, i.e., not directly replicating expressive elements. 
  • In the Bartz vs. Anthropic (Claude AI) case, a US court upheld fair use for AI training on pirated books but acknowledged liability for storage. 
  • In Silverman vs. Meta  (LLaMA AI) case, no market harm was found, but the court emphasized the need for creator compensation frameworks — highlighting evolving legal standards. 
• AI, Creativity & Legal Liability: AI is reshaping copyright norms through AI-generated works (created solely by machines) and AI-assisted works (human-created with AI support).  
  • AI-assisted content is owned by humans, but AI-generated works raise unresolved questions of authorship and ownership.  
  • Legal liability in copyright infringement cases remains unclear, whether on the developer, user, or platform. 

What is the Legal Status of AI-Generated Content in India? 

• Legal Status of AI-Generated Content in India: 
  • Legal Gaps for AI-Generated Content: Indian law currently does not recognize non-human authorship, the Indian Copyright Act, 1957 attributes authorship solely to natural persons, excluding AI and its developers. As a result, AI-generated works without significant human input are not protected.  
    • Additionally, the use of copyrighted material for AI training remains a legal grey area, with no explicit provisions in place.  
    • While Section 52 permits certain uses for "research," its applicability to AI training has not been tested in Indian courts. 
  • Protection for AI-Assisted Works: When a human uses AI as a creative tool, the output may be protected and in such cases, authorship lies with the human, akin to traditional works created using digital tools. 
• Fair Use Provisions under Indian Copyright Law: Section 52 of the Copyright Act, 1957 outlines exceptions to infringement, including: 
  • Private or personal use, including research or education 
  • Criticism or review of any work 
  • Reporting of current events or public lectures 
  • Reproduction for judicial proceedings 
  • Transient or incidental storage during digital transmission or linking 
    • These exceptions form the doctrine of fair dealing, similar to the fair use doctrine in the US. 
• Judicial Interpretation Regarding Use of Copyrighted Material: 
  • Civic Chandran v. Ammini Amma (1996): The Kerala High Court held that parody does not amount to infringement, establishing a 3-factor test: quantity/value of content taken, purpose of use and likelihood of market competition 
  • Eastern Book Company v. D.B. Modak (2008): The Supreme Court (SC) held that raw SC judgments are public domain and not copyrightable, but editorial additions (like headnotes, formatting) by publishers are copyrightable if they show originality through skill and judgment.  
    • The Court rejected the “sweat of the brow” doctrine and adopted the “skill and judgment” test, marking a key judicial interpretation on the originality threshold under Indian copyright law. 
  • India TV v. Yashraj Films (2012): Delhi High Court, in this case expanded fair dealing to cinematographic and musical works, especially after the Copyright (Amendment) Act, 2012, which included exceptions for disabled access and non-commercial public use. 
  • DU Photocopy Case (Oxford v. Rameshwari Photocopy Services, 2016):  The Delhi High Court ruled that photocopying book excerpts for educational use was fair dealing, affirming access to knowledge and public interest as guiding principles. 
• Comparative & Evolving Frameworks: Indian courts increasingly rely on US fair use factors to interpret fair dealing under Section 52 of the Copyright Act, 1957, including the purpose and character of use, nature of the copyrighted work, amount and substantiality, market impact, and transformative character.  
  • However, India lacks a clear definition of “substantial portion”, leaving judicial discretion to determine fair use on a case-by-case basis.  
  • As a TRIPS-compliant nation, India seeks to align with Article 13 of the TRIPS Agreement, which mandates that exceptions to copyright must not conflict with normal exploitation of the work or unreasonably prejudice the rights holder. 
• Policy Developments: A 2025 Commerce Ministry panel is reviewing the Indian Copyright Act, 1957 to address digital and AI-related gaps.  

Comparative Global Approaches to AI-Generated Content 

• US: Copyright is granted only if there’s substantial human creativity (Thaler v. Perlmutter, 2023). Purely AI-generated works are not protected. 
• European Union: The AI Act 2024 mandates transparency of training data. Discussions are ongoing on a sui generis right for AI outputs, though the 2019 Copyright Directive lacks direct provisions. 
• China: Beijing Internet Court delivered a first-of-its-kind judgment in mainland China, recognizing that a picture generated using the AI software Stable Diffusion qualifies as an artwork protected under copyright law.  
  • The court emphasized the “originality” and the intellectual contribution of the human creator, despite the use of AI tools. 
• United Kingdom: Section 9(3) of the Copyright, Designs and Patents Act, 1988 allows copyright for computer-generated works (CGWs) without a human author, assigning authorship to the person making the “necessary arrangements”. However, such works lack moral rights and the provision remains rarely applied due to legal ambiguities and limited judicial interpretation. 

Note 

• In 2021, South Africa became the first country to grant a patent to a machine-generated work, recognizing AI system DABUS as the inventor of a food container design based on fractal geometry.  

Way Forward 

• Legal Modernization & Fair Use Assessment: To address AI-related challenges such as training data usage and algorithmic reproduction, the Indian Copyright Act, 1957 should be updated to include AI-specific provisions.  
  • Additionally, courts can adopt a structured approach to fair use assessment by applying the 4-factor test outlined by the Kerala High Court in Civic Chandran v. C. Ammini Amma (1996), aligning India's framework more closely with US fair use standards. 
• Data Governance & Compliance: Establish clear data usage policies for AI training with oversight mechanisms, audit trails, and mandatory compliance officers in AI firms to ensure adherence to copyright norms and ethical data handling. 
• Balanced Innovation & Rights Protection: Develop a multi-stakeholder regulatory framework to strike a balance between innovation and copyright protection. This includes enabling collective licensing models and ensuring fair compensation for content creators. 
• International Cooperation & Standard Setting: India should proactively engage in global forums like WIPO to shape harmonized copyright rules for AI and contribute to building global ethical and legal standards for generative technologies. 

Drishti Mains Question:

Discuss the key legal and ethical challenges posed by generative AI in relation to copyright laws. Suggest suitable regulatory measures. 

UPSC Civil Services Examination, Previous Year Questions (PYQs) 

Mains

Q. In a globalized world, Intellectual Property Rights assume significance and are a source of litigation. Broadly distinguish between the terms—Copyrights, Patents and Trade Secrets. (2014)