Copyright Protection for AI-Generated Works in India

The advent of Artificial Intelligence (AI) technologies like ChatGPT, Gemini and other generative tools has transformed the way creative content is produced. These AI systems are capable of generating literary works, artworks, music, and even complex research outputs with minimal human intervention. 

However, the rise of such AI-generated content presents unique legal challenges, particularly in the domain of copyright law. India’s existing copyright framework, which is primarily designed around human creativity, does not explicitly address the question of ownership and protection of AI-generated works. 

The Legal Framework: Indian Copyright Act, 1957

The Copyright Act, 1957 is the primary legislation governing copyright protection in India. It recognises authorship as the foundation of copyright and grants exclusive rights to authors for their original works. However, the Act does not expressly define or include AI systems as authors.

• Authorship and Originality: Section 17 of the Act states that the author of a work is the person who creates it. The term ‘person’ is generally understood to mean a natural human being, not a machine or software.
• Computer-Generated Works: The Act was amended in 1994 to include Section 2(d)(vi), which defines “author” for computer-generated works as the person who causes the work to be created. This provision recognises that a computer may generate works autonomously but places authorship with the human agent responsible for such creation.
• Registration Requirements: Application forms for copyright registration (Form XIV) require the claimant’s details such as name, nationality, and address, which presupposes human authorship.

Who Can Be Considered the Author of AI-Generated Works?

There are three potential candidates who could claim authorship over AI-generated content in India:

• AI Systems: Currently, AI programs and systems are not recognised as legal persons under Indian law. They cannot own property, including copyrights. Consequently, AI itself cannot be an author or hold copyright.
• AI Developers: Ownership claims by developers depend on the terms of use and contractual agreements. Most AI platforms like ChatGPT explicitly disclaim ownership over user-generated content. Moreover, just as Microsoft is not an author of a drawing made using Microsoft Paint, AI developers typically are not considered authors of the content generated by users employing their software.
• Users/Prompt-Givers: Section 2(d)(vi) suggests that the person who causes the computer to create a work is the author. However, simply entering a prompt or command may not amount to sufficient human creativity or skill to qualify as authorship.

The Essential Elements for Copyright Protection

In India, for a work to be copyrightable, two core requirements must be met:

• Originality: The work must be original in the sense of reflecting the author’s skill, judgement, and effort. The Supreme Court in Eastern Book Company v. D.B. Modak clarified that originality means a product of skill and judgement that is not purely mechanical or trivial. Mere “sweat of the brow” (effort alone) is insufficient, and the U.S. standard of “modicum of creativity” is deemed too high for Indian law.
• Human Authorship: Copyright law aims to incentivise human creativity. The Constitution’s Copyright Clause reflects this by granting rights to human authors and inventors. Courts have consistently upheld that human intervention and creativity are necessary for copyright protection.

Challenges Posed by AI-Generated Works

AI systems generate content by processing vast amounts of existing data. This raises two key challenges:

• Originality Question: Since AI outputs often recombine pre-existing material, are they truly original? Indian courts require that originality comes from human skill and judgement, which may be lacking if AI generates works independently.
• Authorship and Ownership: When AI acts autonomously or with minimal human input, it becomes difficult to attribute authorship to a human, raising questions about ownership and enforceability of rights.

The ‘Significant Human Input’ Test: A Practical Approach

To bridge the gap between technology and law, a “Significant Human Input” test is increasingly suggested by legal scholars and practitioners. This approach involves two prongs:

• a) Originality through Human Skill and Judgement: The human author’s contribution must go beyond mere prompting or trivial input. It should involve meaningful creative decisions, such as selecting, editing, rearranging, or reshaping AI outputs.
• b) Quantitative and Qualitative Extent of Human Involvement: The degree of human input should be substantial enough that the final work either would not exist or would be fundamentally different without it.

This test respects the traditional copyright framework by insisting on human creativity, while recognising that AI can be a powerful tool in the creative process.

Illustrative Examples

Consider these scenarios:

• Pure AI Generation: An AI writes a story in response to a simple prompt, and the user publishes it without changes. This work lacks significant human input and is unlikely to be protected under copyright law in India.
• Human Editing and Rewriting: A user obtains an AI-generated draft and extensively revises plot, characters, language, and themes. Here, the human involvement crosses the threshold of significant input, and copyright protection may be granted to the human author.
• AI-Assisted Form Corrections: An author writes an article and uses AI tools to correct grammar and improve style. Since the core content is the human’s creation, copyright protection clearly applies.
• Research Enhancement: A legal scholar uses AI to summarise judgments and identify key points but performs original analysis and drafting. The human creativity is the dominant factor, making the work copyrightable.

Fair Use Considerations

Indian copyright law allows limited use of copyrighted material without permission under the “fair use” exceptions for purposes such as criticism, research, or education. While AI training datasets often include copyrighted works, whether such use qualifies as fair use remains an unsettled question. Similarly, whether AI-generated outputs themselves benefit from fair use depends on their nature, purpose, and market impact, requiring case-by-case analysis.

Judicial and Policy Perspectives

Indian courts have yet to directly rule on the copyrightability of purely AI-generated works. However, decisions like Eastern Book Company and Navigators Logistics Ltd. v. Kashif Qureshi emphasise the need for human involvement. Internationally, courts and copyright offices are gradually evolving frameworks to accommodate AI’s role, often reinforcing the primacy of human creativity.

From a policy standpoint, extending copyright protection only to works with significant human input discourages “free riding” on AI-generated content and supports fair remuneration for original human creators. At the same time, it encourages the use of AI as a creative aid rather than a replacement.

Conclusion

AI-generated works are challenging the very foundations of copyright law, which has traditionally protected human creativity and originality. In India, the absence of explicit statutory guidance means that existing principles continue to govern. The key takeaway is that copyright protection will remain reserved for works where significant human creativity and effort shape the final output, even if AI plays a role in the process.

The “Significant Human Input” test strikes a balance between embracing the benefits of AI in creative industries and preserving the core purpose of copyright to incentivise human innovation.