AUTHOR: GUNGUN SHARMA, BMS COLLEGE OF LAW

The creative sector has changed with the advent of Generative Artificial Intelligence (AI). At the touch of a button, platforms like ChatGPT, DALL·E, Bard, and Midjourney can generate poetry, essays, artwork, and even music. Nowadays, human creativity may produce anything in seconds as opposed to hours or days. Although there are many advantages to this technical advancement, it has also led to hitherto unheard-of legal issues. Copyright law is one of the most important issues. The purpose of copyright is to safeguard original works of authorship and guarantee that authors retain control over and profit from their contributions. Does content created by AI, as opposed to a human, qualify for copyright protection? Is it possible to consider AI to be an author? Or does the AI developer or the human user have copyright? The Copyright Act, 1957, which was passed when artificial intelligence was unthinkable, governs copyright law in India. The idea of human authorship is the foundation of the Act. There are uncertainties around AI-generated content because it does not acknowledge computers or non-human beings as creators. This article examines the legal status of AI-generated works under Indian copyright law, including case law, international perspectives, and potential future developments.

Human innovation is at the heart of Indian copyright law. The Copyright Act's Section 2(d) defines "author" in terms of natural persons. For instance: • The author is "the person who creates" a piece of literature or drama. • "The person taking the photograph" is the author of a picture. AI is inherently disqualified as an author by this human-centric criteria. Artificial intelligence (AI) systems are not legal entities, in contrast to companies, which are regarded as "legal persons." They have no legal ability, consciousness, or intention.

This need for human originality has also been reaffirmed by the judiciary. The Supreme Court of India made it clear in Eastern Book Company v. D.B. Modak (2008) that works requiring "skill, judgment, and a minimal degree of creativity" are protected by copyright. The Court decided that simple mechanical work was insufficient for originality in the case, which dealt with the copyrightability of judgments that had been copy-edited. Rather, human intelligence was required. AI-generated works are directly subject to this principle. AI's outputs cannot, by themselves, satisfy the originality criterion required by Indian law because they rely on algorithms, data patterns, and probability models rather than human ingenuity.

Humans who make major contributions to the process of innovation can claim ownership, even though AI cannot be acknowledged as an author. The level of human participation is crucial. A work would probably not be protected by copyright if a user merely enters a generic command, like "write a poem on monsoon," and accepts the AI's unedited result. Due to the small amount of human involvement, it would not meet the originality requirement. However, the final product becomes a collaborative effort if the human adds personal creative features, reorganizes the framework, edits the AI's draft, or gives explicit directions. In this case, human "skill and judgment" are crucial, which qualifies the work for copyright protection.

In particular, the D.B. Modak standard is pertinent. The originality requirement may be met by significantly impacting AI outputs by human interaction, much like copy-editing required decisions and mental work. The function of AI developers must also be made clear. Users' rights to outputs are expressly stated in the terms of service of the majority of platforms, such as Google (Bard) and OpenAI (ChatGPT). Usually, developers deny ownership. They play a similar position as a tool provider, much like Adobe does with Photoshop or Microsoft with MS Paint. Developers are therefore unable to assert copyright over products made with their tools.

Under Indian law, works created entirely by artificial intelligence (AI) without human intervention are not protected by copyright. This is due to:

 1. They do not meet Section 2(d)'s criterion for human authorship. 2. They do not meet the "skill and judgment" originality standard established in D.B. Modak. This strategy guarantees that human creation, not machine production, will continue to be rewarded by copyright law. AI can still be a very effective assistive technology, though. For instance, a researcher may use AI to compile data, a designer may use AI to create initial drafts and subsequently revise them, or an author may use AI to improve language or grammar. The finished work in each of these situations showcases human innovation and might be covered by copyright. . Scholars propose a "Significant Human Input Test" to differentiate between works that are protected and those that are not. Only when the human contribution is significant enough that the work would not exist in its current form without human involvement would copyright protection be applicable under this paradigm. This criterion ensures that copyright is anchored in creativity rather than automation by striking a balance between innovation and legal certainty.

The use of copyrighted materials to train AI models is a particularly contentious topic. Books, scholarly articles, movies, songs, and photos, many of which are copyright protected, are frequently included in the large datasets used to train AI. Unauthorized reproduction of such works may be considered infringement under Indian law. Limited "fair dealing" exceptions are granted for private research, criticism, review, and educational usage under Section 52 of the Copyright Act. These exclusions do not, however, apply to extensive, commercial AI training.

This problem has led to international litigation. For instance: 

1) OpenAI has been sued by writers in the US for utilizing their copyrighted novels as training datasets. 2) Artists in the UK have criticized AI firms for using copyrighted music to train models without permission. Similar lawsuits have not yet occurred in India, but disagreements are unavoidable as AI adoption increases. Both the AI company and the user deploying the content may be held liable if the AI-generated content closely resembles copyrighted works.

The Indian judiciary is currently starting to look into copyright concerns pertaining to AI. ANI Media v. OpenAI is a noteworthy lawsuit that is presently proceeding before the Delhi High Court. This is the first significant case in India to examine the consequences of generative AI for copyright. Future regulations will probably be influenced by the Court's observations.

Divergent methods are also revealed by comparative insights: • At first, the Copyright Office in the US denied copyright for works that were solely produced by artificial intelligence.

 • In the UK, computer-generated works are permitted under Section 9(3) of the Copyright, Designs and Patents Act, 1988, where the "person who made the arrangements" is regarded as the author. Nevertheless, in the Zarya of the Dawn case, it acknowledged limited protection where a human assembled AI-generated images into a graphic novel. There is a legal loophole in India since there is no such clause. India may require legislation revision in light of these advancements in order to define the legal status of AI-generated and AI-assisted works.

Several Indian instances offer guiding concepts even though they are not specifically related to AI: 1)  In R.G. Anand v. Deluxe Films (1978), the Supreme Court stressed that ideas are not protected by copyright; only expression is. Infringement may occur if AI-generated products replicate preexisting expressions. 2) In the 2018 case of Navigators Logistics v. Kashif Qureshi, the Delhi High Court reaffirmed that originality necessitates a certain level of inventiveness. This idea holds true for AI-assisted projects as well. 3) Gramophone Company of India v. Super Cassettes (2010): The Court emphasized the need to safeguard the economic and moral rights of writers, which could be jeopardized by unapproved AI training. These examples demonstrate how the Indian judiciary values human ingenuity and the defence of preexisting rights, a position that is probably going to hold for issues involving artificial intelligence.

The legal system in India must change to meet the problems posed by AI. Among the potential strategies are: 1. Introducing the "Human Input" Test to guarantee that only works involving a great deal of human inventiveness are protected by copyright. 2. Establishing a New Category for AI Works: Like the UK, India might acknowledge works produced by AI while giving credit to the human who developed the arrangements. 3. Regulating AI Training: Possibly using licensing frameworks, this involves establishing explicit guidelines for the use of copyrighted works in training datasets. 4. Judicial Guidance: Courts ought to establish uniform guidelines that strike a balance between creativity and the defense of creators' rights. Such changes would guarantee that copyright law keeps up with technology advancements and provide creators, AI firms, and users peace of mind.

For Indian copyright law, generative AI poses both opportunities and challenges. AI is not allowed to create content since the Copyright Act of 1957 clearly placed human writing at the forefront. Therefore, works created solely by AI are not protected. However, AI-assisted creations that incorporate a great deal of human ingenuity may be eligible for copyright, guaranteeing that human talent and discernment will always be valued. However, there are serious infringement issues with the unapproved usage of copyrighted content for AI training. Legislators must address this by enacting clear legislation provisions or licensing regimes.

The legislative and courts must cooperate to find a balance as India finds itself at the intersection of legal ambiguity and technological advancement. The future depends on maintaining human authorship as the cornerstone of copyright law while acknowledging AI as a tool of creativity rather than a creator in and of itself.