AI and Copyright in India: Who Owns What AI Creates?

Can Artificial Intelligence Be an Author?

Artificial Intelligence is increasingly becoming part of the creative and commercial process.

From generating images, drafting marketing content and producing music to assisting with software, advertising, publishing and design, generative AI can now produce sophisticated outputs with limited human intervention.

But an important legal question follows:

If AI creates a work, who is the author?

A recent decision of the Indian Copyright Office has brought this question into sharp focus.

In an order dated 31 August 2026, the Copyright Office considered an application concerning an artwork titled A Recent Entrance to Paradise, generated by Stephen Thaler's artificial intelligence system, DABUS (Device for the Autonomous Bootstrapping of Unified Sentience).

The Copyright Office drew an important distinction between originality and legal authorship. While the AI-generated artwork was found to satisfy the originality requirement, DABUS itself could not be recognised as the legal author under the existing framework of Indian copyright law.

The decision provides an important starting point for understanding how Indian copyright law currently approaches AI-generated works.

1. What Happened in the DABUS Case?

The dispute concerned an artistic work titled A Recent Entrance to Paradise.

Stephen Thaler sought copyright registration in India and identified DABUS, his AI system, as the author of the work, while claiming copyright ownership himself.

The application was filed in 2022 and subsequently raised questions before the Copyright Office concerning the identity and legal status of the claimed author.

The central question was not simply whether the artwork was created using artificial intelligence.

Rather, the Copyright Office had to consider whether an AI system could legally be recognised as the author of a copyrightable work under Indian law.

The Registrar of Copyrights ultimately held that DABUS could not be recognised as the author because it is neither a natural person nor a juristic person capable of holding legal rights and obligations.

The Office also considered the fact that Thaler had created and configured the AI system and initiated the process that resulted in the artwork.

However, the application continued to be maintained on the basis that DABUS itself was the author, despite an opportunity being provided to amend the authorship particulars.

The application was therefore rejected on that basis.

Importantly, the rejection does not mean that AI-generated works are categorically incapable of satisfying copyright requirements.

That distinction is critical.

2. Originality Is Not the Same as Authorship

The most significant aspect of the decision is the distinction between two separate concepts:

Originality

Copyright law protects original works.

Under Section 13 of the Copyright Act, 1957, copyright subsists in original literary, dramatic, musical and artistic works, subject to the statutory framework.

The Copyright Act does not provide a simple definition of "originality". Indian courts have developed the applicable principles through case law, including the Supreme Court's decision in Eastern Book Company v. D.B. Modak.

Broadly, originality requires more than mere mechanical reproduction. A work must contain the requisite degree of independent expression or creativity.

Authorship

Authorship is a separate question.

Even if a work is sufficiently original to qualify for copyright protection, the law must still identify the person who can legally be recognised as its author.

This is where AI presents a difficult question.

An AI system may generate the final expression.

But an AI system is not presently treated under Indian copyright law as a legal person capable of holding copyright in its own name.

Therefore:

An AI-generated work may satisfy the originality requirement without the AI system itself becoming its legal author.

That distinction is likely to become increasingly important as generative AI becomes more autonomous.

3. What Does Indian Copyright Law Say About Computer-Generated Works?

Interestingly, Indian copyright legislation already contains language specifically dealing with computer-generated works.

Section 2(d)(vi) of the Copyright Act defines the "author", in relation to a literary, dramatic, musical or artistic work which is computer-generated, as:

"the person who causes the work to be created."

This provision is particularly important in the AI context.

It indicates that Indian copyright law does not necessarily require the person to physically create every element of a computer-generated work.

Instead, the legislation focuses on the person who causes the work to be created.

The DABUS decision therefore raises a broader question:

When an AI system generates an output with little or no human intervention at the final stage, who is the person who legally "causes" the work to be created?

The answer may depend heavily on the facts.

4. Is the Person Using AI Automatically the Author?

Not necessarily.

This is where businesses and creators need to exercise caution.

Consider two different situations.

Scenario 1: AI-Assisted Creation

A graphic designer develops a concept, selects the composition, provides detailed instructions to an AI tool, repeatedly modifies the prompts, selects specific outputs and substantially edits the final image.

Here, the human contribution may be significant.

The AI may function primarily as a tool within a broader human creative process.

Scenario 2: Predominantly Autonomous Generation

A person provides minimal input to an AI system and the system independently generates the final work, with little meaningful human creative intervention.

The authorship analysis becomes considerably more complicated.

Simply paying for an AI subscription or pressing a "generate" button does not necessarily answer the legal question of authorship.

The degree, nature and significance of human involvement may become increasingly important.

5. Does This Mean AI-Generated Content Has No Copyright in India?

No — that conclusion would be too broad.

The DABUS decision should not be interpreted as establishing that every work generated using AI is automatically outside copyright protection.

The Copyright Office's reasoning is more nuanced.

The artwork in question was considered capable of satisfying the originality requirement.

The difficulty arose because the application sought to recognise DABUS itself as the author.

Therefore, the decision highlights a distinction between:

Whether a work can be sufficiently original and Who can legally be recognised as its author.

This distinction will be particularly relevant to AI-assisted creative work.

The legal position may also evolve as Indian courts, the Copyright Office and the legislature encounter more sophisticated forms of generative AI.

6. Why Does Authorship Matter?

Authorship is not merely a label.

It has significant consequences for ownership and commercial exploitation.

Section 17 of the Copyright Act generally provides that the author is the first owner of copyright, subject to statutory exceptions.

This means that identifying the author can directly affect questions such as:

• Who owns the copyright?

• Who can license the work?

• Who can assign the copyright?

• Who can enforce infringement claims?

• Who can commercially exploit the work?

• Who can grant rights to third parties?

• Who bears responsibility for representations concerning ownership?

  
  

An AI system cannot presently perform these legal functions in its own name.

This is why the question of AI authorship is also a question of commercial ownership and enforceability.

7. AI-Assisted Works vs. AI-Generated Works

For businesses, one of the most useful distinctions is between AI-assisted and AI-generated content.

AI-Assisted Work

AI is used as a tool, but the human remains substantially involved in the creative process.

Examples may include:

• A marketing team using AI to generate initial copy and then substantially rewriting it.

• A designer using AI to generate concepts and manually creating the final artwork.

• A filmmaker using AI for storyboarding while retaining creative control over the final production.

• A lawyer using AI to assist with drafting while independently reviewing, revising and finalising the document.

In such situations, documenting the human contribution can become important.

AI-Generated Work

The AI system generates the substantive output with limited human creative intervention.

Examples may include:

• An image generated from a short prompt with no subsequent modification.

• AI-generated music produced without meaningful human composition or editing.

• Entirely machine-generated text published without substantive human intervention.

These situations raise more difficult questions about authorship, ownership and the scope of copyright protection.

The legal analysis should therefore focus on how the work was actually created, rather than simply whether an AI tool was used.

8. What Does This Mean for Businesses?

The implications extend far beyond artists and content creators.

Businesses increasingly use generative AI for:

• Advertising campaigns

• Website content

• Product descriptions

• Social media posts

• Graphic design

• Presentations

• Software development

• Marketing materials

• Training materials

• Video production

• Music and audio

• Internal documentation

• Customer communications

  
  

The legal risks can include both ownership risk and third-party intellectual property risk.

For example, a company may assume that because it paid for an AI service, it automatically owns all rights in every output.

That assumption may not always be correct.

The contractual terms of the AI provider, the nature of the output, the level of human contribution and applicable intellectual property law must all be considered.

9. Contractual Protection Becomes Increasingly Important

Businesses using AI should consider addressing AI-generated content expressly in their commercial agreements.

Depending on the transaction, contracts may address:

• Ownership

  Who owns AI-assisted deliverables?

• Human Contribution

  Who is responsible for ensuring that sufficient human creative contribution exists where required?

• Third-Party Rights

  Who bears responsibility if an AI-generated output allegedly infringes a third party's intellectual property?

• Warranties and Representations

  Can the service provider represent that the content is appropriately licensed or cleared for the intended commercial use?

• Indemnities

  What happens if a third party brings an intellectual property claim?

• Confidentiality

  Can confidential business information or personal data be entered into the AI system?

• Data Usage

  How does the AI provider use prompts, uploaded material and generated outputs?

• Regulatory Compliance

  Does the use of AI comply with applicable privacy, intellectual property and sector-specific requirements?

  
  

These provisions can be particularly important for agencies, technology companies, media organisations, startups and businesses outsourcing creative work.

10. AI Creates a New Layer of IP Due Diligence

Traditional intellectual property due diligence generally asks questions such as:

• Who created the work?

• Who owns it?

• Was it assigned?

• Are there third-party licences?

• Are there infringement claims?

• Are registrations available?

  
  

AI introduces additional questions.

Businesses may also need to ask:

• Was AI used to create the work?

• Which AI system was used?

• What human contribution was made?

• What were the relevant prompts or inputs?

• Were third-party materials used?

• What do the AI provider's terms say about ownership?

• Was confidential information provided to the AI system?

• Can the company demonstrate its rights in the final output?

These questions can become particularly important during investments, acquisitions, licensing transactions and commercial disputes.

11. What Should Businesses Do Now?

Businesses do not necessarily need to stop using generative AI.

Instead, they should develop a structured AI governance approach.

Some practical measures include:

• Maintain an AI-use policy

  Define which AI tools employees can use and for what purposes.

• Document human involvement

  For important creative works, maintain records of the human creative process, edits and decisions.

• Review AI-provider terms

  Understand the provider's provisions concerning ownership, licensing, training, data usage and commercial exploitation.

• Protect confidential information

  Employees should not upload confidential business information, client information or sensitive material into AI systems without appropriate safeguards.

• Conduct IP checks

  AI-generated content should not automatically be assumed to be free from third-party intellectual property concerns.

• Include AI clauses in contracts

  Client agreements, vendor agreements, employment arrangements and creative-services contracts may need AI-specific provisions.

• Establish approval procedures

  High-value or externally published AI-generated material should be reviewed before commercial use.

12. The Bigger Legal Question: Who Is the Creator?

The DABUS decision also raises a philosophical question with very practical legal consequences.

Traditional copyright law was developed around human creativity.

Generative AI challenges the assumption that the person who ultimately produces the visible or audible expression must necessarily be the person who directly created every element of it.

But copyright law does not operate solely on the basis of technological capability.

It operates through legal concepts such as:

author → ownership → rights → licensing → enforcement.

Artificial intelligence may be capable of generating sophisticated expression.

But the legal system still needs a legally recognisable rights-holder.

The Indian Copyright Office's recent decision therefore represents an important moment in the development of India's approach to AI and intellectual property.

13. What the DABUS Decision Does — and Does Not — Decide

It is important not to overstate the ruling.

The decision does not establish that:

• Every AI-generated work is automatically copyright-protected.

• Every person who uses AI automatically becomes the copyright owner.

• Human involvement is irrelevant.

• AI-assisted works can never receive copyright protection.

• Indian copyright law has permanently settled every question surrounding generative AI.

  
  

Instead, the decision provides an important indication of the current position:

AI itself cannot presently be recognised as the legal author under the existing Indian copyright framework.

At the same time, the fact that an AI system generated a work does not, by itself, necessarily prevent the work from satisfying the originality requirement.

The precise legal position will continue to develop through future administrative decisions, judicial interpretation and potentially legislative reform.

14. What Should Creators Keep in Mind?

For individual creators, the practical takeaway is simple:

Keep records of your creative process.

If you use AI in producing commercially important content, retain evidence of:

• Your original concept

• Prompts and instructions

• Drafts

• Editing history

• Human modifications

• Final creative decisions

• AI tools used

• Relevant licence or subscription terms

This documentation may become valuable if ownership or authorship is subsequently challenged.

15. The Road Ahead for AI and Copyright in India

The DABUS decision is unlikely to be the final word on AI and copyright.

Generative AI is developing rapidly.

Systems are becoming more capable of producing images, music, text, video and software with increasingly limited human input.

This creates a regulatory challenge:

How should copyright law balance technological innovation with the fundamental concepts of human authorship, ownership and accountability?

India's existing statutory framework already contains a provision dealing with computer-generated works.

However, the rise of modern generative AI presents questions that could not have been fully anticipated when the legislation was framed.

Future disputes may require courts and policymakers to consider:

• What constitutes sufficient human creative contribution?

• How should AI-assisted works be distinguished from autonomous outputs?

• Who owns AI-generated commercial content?

• How should AI training and input data affect copyright analysis?

• How should infringement be assessed where AI generates similar outputs?

• Should the law evolve to expressly address generative AI?

• What contractual standards should businesses adopt?

  
  

These questions will become increasingly relevant as AI moves from an experimental technology to an ordinary business tool.

Conclusion

The DABUS decision highlights a fundamental principle in India's emerging AI-copyright landscape:

AI may be capable of creating an original output, but that does not mean AI itself can presently be recognised as the legal author.

For businesses and creators, the lesson is not to avoid AI.

It is to use AI deliberately, transparently and with appropriate legal safeguards.

The more significant the commercial value of an AI-assisted work, the more important it becomes to understand:

Who created it?Who contributed to it?Who owns it?What rights exist?What third-party risks arise?And can those rights actually be enforced?

As artificial intelligence becomes increasingly embedded in creative and commercial processes, intellectual property strategy must evolve alongside the technology.

Innovation creates opportunity. Legal clarity makes it sustainable.

How Valkas Legal Can Help

At Valkas Legal, we advise businesses, startups, entrepreneurs, technology companies and creators on legal issues arising at the intersection of technology, intellectual property, media and commercial law.

Our advisory approach can assist businesses with:

• AI governance and advisory

• Intellectual property strategy

• Commercial contracts

• Technology agreements

• Data privacy and digital compliance

• Media and entertainment advisory

• IP-related risk management

• AI-related contractual protections

• Commercial and regulatory advisory

  
  

For businesses adopting generative AI, legal strategy should be considered before a dispute arises — not after.

                                                                                                                          -Valkas Legal