Introduction
Nowadays, we can see our growing dependence on artificial intelligence (AI) and how often we use it in our day-to-day lives. We all know that using AI has many advantages, such as generating creative photos, ideas, videos, music, etc.
But on the other side, or hidden side, it has some strong harmful effects, such as heavily affecting the creative thinking of human beings, causing cases of data breach, and also impacting the authors, artists, and businesses in their work.
Growing Concerns Over AI-Generated Content
There is a question that arises: if artificial intelligence creates a work, then who owns it? Because people are taking ownership of work created by AI, which is not of their own.
Here comes the Copyright Law, which requires legal and policy reforms to clarify the ownership and protect the rights of creators.
Scope of the Article
This article discusses in detail the fundamentals of copyright law, ownership of AI-generated content, global developments, and its position under Indian law.
| Key Topics Covered | Description |
|---|---|
| Copyright Law Fundamentals | Understanding the basic principles of copyright protection. |
| AI-Generated Content Ownership | Examining who owns content created using artificial intelligence. |
| Global Legal Developments | Analysing international approaches toward AI and copyright law. |
| Indian Legal Position | Discussing the status of AI-generated works under Indian law. |
- Artificial intelligence is increasingly becoming part of everyday life.
- AI tools can generate photos, videos, music, and creative ideas.
- AI also raises concerns related to creativity, ownership, and data privacy.
- Copyright law plays an important role in protecting creators’ rights.
Understanding the Word ‘Generative AI’
‘Generative AI’ refers to artificial intelligence systems that generate content, such as photos, videos, and text, based on user prompts. These systems use machine learning models, often built on large language models (LLMs) like deep learning. The term “generative” highlights their capability to autonomously create new outputs from provided data or instructions. [1]
For the past ten years, artificial intelligence (AI) has been a popular technical issue. However, generative AI—more especially, the introduction of ChatGPT in 2022—has brought AI to the attention of people all over the world and sparked an unprecedented wave of AI research and acceptance.
Businesses are moving forward, investigating how the technology may enhance their internal workflows and enhance their goods and services, despite the fact that generative AI offers huge productivity benefits for both individuals and organisations.
A third of companies already routinely use generative AI in at least one business activity, according to data from the management consulting firm McKinsey. More than 80% of businesses will have implemented generative AI applications or leveraged generative AI application programming interfaces (APIs) by 2026, according to industry analyst Gartner. [2]
Training Data and Copyright Concerns
We have a lot of information, like books and articles, photos, and code, that we use to teach the artificial intelligence systems. Some of this information might be owned by someone. This makes us think about two things.
- Can we say that the things created by artificial intelligence systems are owned by someone because of copyright?
- Is it all right to use information that is owned by someone to teach artificial intelligence models?
These questions are very important when we talk about the problems with copyright that we are having now. The artificial intelligence system is changing the way we think about copyright.
How Does the Artificial Intelligence System Affect the Rules About Copyright in India?
In India, the artificial intelligence system has to follow the rules about copyright because the Copyright Act of 1957 says that it protects the work of the person who created it. The person who created it is defined in Section 2(d) of the Copyright Act of 1957 [3].
The section says the following:
| Clause | Meaning Under Section 2(d): |
|---|---|
| (i) | In relation to a literary or dramatic work, the author of the work; |
| (ii) | In relation to a musical work, the composer |
| (iii) | In relation to an artistic work other than a photograph, the artist |
| (iv) | In relation to a photograph, the person taking the photograph; |
| (v) | In relation to a cinematograph film or sound recording, the producer and |
| (vi) | In relation to any literary, dramatic, musical or artistic work which is computer-generated, the person who causes the work to be created also [4] |
Requirements for Copyright Protection
- It must be original or unique, and
- It must be in tangible form, i.e., other people could see, hear, and experience it.
Supreme Court View on Originality
The Supreme Court of India ruled in Eastern Book Company v. D.B. Modak that a work does not have to be brand-new to be protected by copyright.
What counts is that it “originates from the author” and requires a certain amount of skill and judgement; for instance, selecting, organising, revising, or assembling information in a novel way can all contribute to originality. [5]
The First Challenge: Ownership of AI-Generated Outputs
Whether a machine can be identified as an author is the most challenging question posed by generative AI. The majority of legal systems provide a negative response to this query. The idea of human creation is the foundation of copyright law. Legal individuality, purpose, and consciousness are non-existent in machines. Instead of being entities with rights, they are instruments. [6]
The US Copyright Office has emphasised time and time again that only works made by humans are protected by copyright. Applications with only AI-generated photos have been rejected unless they show a significant amount of human creativity. [7]
This approach suggests that AI-generated content may fall into one of three categories:
- User-Owned Works – where the user exercises significant creative control.
- Developer-Owned Works – where ownership is claimed by the software provider.
- Public Domain Outputs – where no copyright exists because there is no human author.
Among these, the first and third approaches are the most widely discussed.
Categories of AI-Generated Content
| Category | Description | Key Legal Concern |
|---|---|---|
| User-Owned Works | Content where users exercise significant creative control. | Determining the level of human creativity required for copyright protection. |
| Developer-Owned Works | Ownership claimed by the software or AI provider. | Whether AI developers can legally claim authorship rights. |
| Public Domain Outputs | Outputs without human authorship. | Whether AI-generated works should remain free for public use. |
The Second Challenge: Use of Copyrighted Works for Training AI
Think of a time that you have copyrighted an audio to protect it from unauthorised use, but the next day, what you see is that other persons are using it without your authority, with the help of AI. This is our second and crucial challenge of using AI in regard to copyright law. There is a lot of copyrighted content available on the internet, & AI has built its database on such content available online. With this, AI provides us with original or similar kinds of copyrighted content, which harms the real creator of that content. We have laws of copyrighting the work, but not related to the use of that work by generative AI models. [8]
Although the technological context differed from modern AI systems, the controversy similarly revolved around whether copying copyrighted material for technological processing constituted infringement or a permissible use. This clearly showcases how emerging technologies repeatedly test the limits of traditional copyright doctrines.
Andersen v. Stability AI: Case Analysis
In Andersen v. Stability AI, for instance, the court determined that the plaintiffs’ claims that Stability obtained copies of the plaintiffs’ copyrighted works, used those works to train Stable Diffusion, and then stored or incorporated the training images derived from their works into Stable Diffusion as compressed copies were sufficient to plead direct infringement.
The plaintiffs claimed that Stability AI, Midjourney, and DeviantArt scraped billions of images from the internet, including the plaintiffs’ copyright. The court determined that the plaintiffs’ claims that Stability obtained copies of their copyrighted works, trained Stable Diffusion using those works, and then stored or integrated the training images derived from their works into Stable Diffusion as compressed copies were adequate to establish direct infringement. [9]
Major Copyright Concerns in AI Training
- unauthorised scraping of copyrighted material from the internet.
- Absence of consent from original creators.
- Generation of substantially similar copyrighted outputs.
- Lack of clear legal provisions governing AI training datasets.
- Balancing innovation with protection of creators’ rights.
Impact of Generative AI on Copyright Law
| Issue | Impact on Copyright Law |
|---|---|
| AI Authorship | Challenges the traditional requirement of human authorship. |
| Training on Copyrighted Data | Raises concerns about infringement and fair use. |
| Ownership of Outputs | Creates uncertainty regarding rights over AI-generated works. |
| Creator Protection | Demands stronger legal safeguards for original creators. |
Recent Global Developments In AI And Copyright Law
Lawmakers are working on laws to deal with artificial intelligence and copyright concerns. For example,
- The Generative AI Copyright Disclosure Act of 2024 was introduced in the U.S. Congress on April 9, 2024. This law would make companies that create intelligence models tell people what data they used to train their systems. This will make things more transparent. It will give people who own copyrights control over their work. [10]
- The No AI FRAUD Act was introduced in 2024 to stop intelligence from being used to pretend to be someone without their permission. This could affect the entertainment industry, where artificial intelligence is being used to create performances.
- The European Union AI Act is a proposed law that wants to create rules for intelligence technologies. It uses a risk-based approach to regulate intelligence systems, which means it looks at how risky they are and makes rules based on that.
- The AI Act says that artificial intelligence developers have to keep records of the data they use to train their systems. This makes sure that artificial intelligence development is transparent and accountable. It might also lead to copyright protections.
Position Under Indian Law On AI Copyright
In India, there was a case in 2020 about an artificial intelligence system called RAGHAV. RAGHAV tried to get a copyright for a painting called ‘Suryast’. The painting was not allowed to have a copyright at first because it did not have an author.
Then it was allowed to have a copyright when a human was named as a co-author with RAGHAV. Then a notice was sent to ask for clarification on what RAGHAV’s status is [11].
This showed that there is confusion about whether artificial intelligence can be considered an artist under the copyright law.
Key Issues In The RAGHAV Case
| Issue | Details |
|---|---|
| Artwork | ‘Suryast’ |
| AI System | RAGHAV |
| Main Concern | Whether artificial intelligence can be treated as an author |
| Legal Confusion | Status of AI under copyright law |
Recommendations For AI And Copyright Regulation
To deal with the copyright problems caused by intelligence, lawmakers should make it clear who owns artificial intelligence-created works. Copyright protection should be given when a human has made a creative contribution to the final work.
In India, the Copyright Act of 1957 should be changed. This change would explain how the concept of “the person who causes the work to be created” applies to intelligence systems.
Recommended Legal Reforms For AI Copyright
- Create clear ownership rules for artificial intelligence-created works.
- Give copyright protection where there is meaningful human creativity.
- Amend the Copyright Act of 1957 to address intelligence systems specifically.
- Regulate the use of copyrighted materials for training intelligence models.
- Require artificial intelligence developers to obtain licences before using protected content.
- Ensure authors and publishers receive fair compensation.
- Make companies disclose the kinds of works used for training their artificial intelligence systems.
- Allow creators to opt out of having their work used in artificial intelligence training.
- Maintain a balance between innovation and creators’ rights.
There should be a set of rules to regulate the use of copyrighted works to train intelligence models. Artificial intelligence developers should get permission to use protected materials by getting a licence. This license should make sure that authors and publishers get paid.
Companies should have to tell people what kinds of work they use to train their artificial intelligence models. Creators should be able to choose not to let their work be used.
Any changes to the law should balance the need to promote innovation and the need to protect the rights of creators. This balance will let both artificial intelligence and the creative industries develop in a way.
Conclusion
Artificial intelligence has changed the way we think about who creates a work, what is original, and who owns a work. When machines help create something, it is hard for the law to determine who should own it.
The problem is made worse by the fact that artificial intelligence models often use copyrighted works without permission. While artificial intelligence offers opportunities for innovation and creativity, it also raises concerns for authors, artists, and publishers.
A fair and balanced approach to the law is necessary to protect creators’ rights and to encourage progress. In India, the Copyright Act of 1957 is a starting point. More clarity is needed from the courts and lawmakers.
The future of copyright law depends on its ability to adapt to a world where humans and machines work together to create things.
References
- https://www.ibm.com/think/topics/generative-ai
- https://www.gartner.com/en/newsroom/press-releases/2023-10-11-gartner-says-more-than-80-percent-of-enterprises-will-have-used-generative-ai-apis-or-deployed-generative-ai-enabled-applications-by-2026
- https://www.khuranaandkhurana.com/ownership-of-ai-generated-content-a-deep-dive-into-copyright-law-in-india
- https://indiankanoon.org/doc/1655540/
- https://markshield.in/understanding-basics-of-copyright-law-in-india/
- https://www.khuranaandkhurana.com/ownership-of-ai-generated-content-a-deep-dive-into-copyright-law-in-india
- https://www.insidetechlaw.com/blog/2024/05/generative-ai-how-it-works-content-ownership-and-copyrights
- https://en.wikipedia.org/wiki/Artificial_intelligence_and_copyright#Training_AI_with_copyrighted_data
- https://www.loeb.com/en/insights/publications/2023/11/andersen-v-stability-ai-ltd?utm_source=chatgpt.com
- https://sites.usc.edu/iptls/2025/02/04/ai-copyright-and-the-law-the-ongoing-battle-over-intellectual-property-rights/
- https://www.barandbench.com/view-point/intersection-intellectual-property-rights-ai-generated-works-part-i
