Artificial Intelligence and Copyright: Need for Legal Reform in India

I. Introduction.

In this era of technological revolution, Artificial Intelligence (AI) is no longer just simplifying human tasks; it has aggressively stepped into the realm of creativity. From composing musical symphonies to painting digital artwork and drafting literary pieces, AI is rapidly expanding its footprint into domains once considered the exclusive monopoly of the human mind.

As the influx of AI-generated content continues to surge, a fierce debate has emerged within legal circles. The core dilemma revolves around a fundamental question: who deserves the actual copyright ownership of these creations? Should it belong to the machine that generated the work, or the human who provided the prompts and instructions? This ambiguity has severely disrupted the traditional frameworks of Intellectual Property Rights (IPR).

The primary hurdle here is that the Indian Copyright Act, 1957, fundamentally extends legal protection only to a “human author.” Under the current statutory regime, a work is deemed eligible for copyright only if it involves human intellect, labor, and skill. Because the law lacks any explicit provision to recognize a non-human entity or software as an “author,” it has created a massive legal vacuum and profound uncertainty regarding AI-generated works.

This article argues that India’s existing copyright framework is entirely inadequate to address the novel and complex challenges posed by modern Artificial Intelligence. While technology has advanced at a breakneck pace, our legal regulations have lagged significantly behind, necessitating immediate legislative intervention and reform.

To systematically unpack this legal conundrum, the subsequent sections of this paper are organized as follows: Part I scrutinizes the evolving legal definition of “Authorship” in the age of AI. Part II evaluates how the standards of “Originality” should be measured concerning AI-generated outputs. Part III analyzes the judicial interpretation and stance taken by Indian courts on related intellectual property matters so far. Finally, Part IV concludes by proposing pragmatic legal reforms to bridge the current statutory gaps and make the law future-ready.

II. Meaning and Growth of Artificial Intelligence

To accurately analyze this escalating legal crisis, we must first look past the marketing buzzwords and understand what Artificial Intelligence actually is from a technical and operational perspective. In simple terms, AI refers to a cluster of advanced software systems, deep learning models, and complex data-processing algorithms that mimic human cognitive intelligence to perform tasks like experiential learning, logical reasoning, complex decision-making, pattern recognition, and creative generation.

If we look back at the historical growth of this technology, we can see a very clear split between what AI used to be and what it has become today. In the initial phases of computer science, technology operated strictly under the “assistive tool” model. When an early computer software program acted as a calculator, a basic spell-checker, or a simple design editor, humans did 100% of the creative heavy lifting. The computer was just a high-tech pen or a faster typewriter; it made zero creative or selective choices on its own. It merely executed the manual, direct commands of the human controller.

However, with the sudden advent of advanced Machine Learning (ML), Natural Language Processing (NLP), and Deep Neural Networks, we have fully entered the disruptive era of “Generative AI.” Modern platforms like ChatGPT, Midjourney, Claude, and Stable Diffusion can now independently produce highly sophisticated outputs based on extremely minimal human prompts.

The critical turning point is this: AI is no longer just a passive assistant; it has effectively transformed into an independent creator. When a generative AI tool creates a highly complex painting, designs a commercial logo, or writes a flawless academic essay, it is not simply cutting, pasting, or rearranging existing files from the internet. Instead, it is generating a brand-new, unique expression by analyzing billions of historical data patterns and calculating mathematical probabilities. This shifts the entire philosophical paradigm of creativity. The traditional line separating human expression from automated machine output has become completely blurred, causing severe waves of anxiety in copyright law.

III. Understanding Copyright Law in India

Copyright is a negative legal right granted by the state to creators to protect their original literary, dramatic, musical, and artistic works from unauthorized commercial exploitation. In our domestic jurisdiction, the primary legislation governing this domain is the Indian Copyright Act, 1957. The fundamental philosophy behind copyright protection is to incentivize human creativity. By granting exclusive economic and moral rights to creators, the law ensures that individuals can safely profit from their intellectual labor, thereby enriching the nation’s cultural and academic repository without fear of unauthorized duplication or theft.

To grasp exactly how this law interacts with modern Artificial Intelligence, we must examine three critical, interconnected sections of the Act:

Section 2(d): This section functions as the statutory gatekeeper by strictly defining the term “author.” It specifies who the author is for different types of creative works (e.g., the author of a book, the composer of a musical track, the photographer of a picture).

Section 13: This provision details the specific categories of works in which copyright can legally subsist, explicitly stating that protection is only available for works that are “original.”

Section 17: This section sets up the foundational baseline of ownership by declaring that the author of a work is generally the “first owner” of the copyright.

The core takeaway from reading these statutory provisions together is absolutely clear: Indian copyright law was built by humans, explicitly for humans. The entire statutory architecture assumes that a living, breathing person with consciousness and intent is behind every creative stroke. When the Act was drafted in the mid-20th century, machines were completely mechanical and incapable of independent thought, meaning the lawmakers simply never anticipated a world where a software program could generate art autonomously.

IV. Ownership Problem Under Section 2(d) and Section 17

This brings us face-to-face with the single biggest statutory roadblock in our current intellectual property framework. Section 2(d) defines an author, and for computer-generated works, Section 2(d)(vi) states that the author is “the person who causes the work to be created.” While this specific clause sounds beautifully flexible and broad on paper, it creates an intense, chaotic ownership confusion the moment you try to apply it to a modern generative AI ecosystem. If an AI tool creates an award-winning piece of digital art, who exactly is the “person” who caused that work to be created?

If we dissect the reality of how an AI output comes to life, we are left with four distinct possibilities, and unfortunately, every single one of them contains deep logical and legal flaws:

The AI Software Developer: One argument claims that the programmer who wrote the AI’s original source code should get the copyright. However, the developer has absolutely no idea what an end-user will type into the system years down the line. The developer did not contribute a single thought or creative choice to the specific output, so claiming ownership over millions of random user outputs across the globe is unfair and commercially unfeasible.

The End-User: The second argument suggests that the person who typed the prompt should be the legal author. But can typing a simple, lazy five-word prompt like “draw a futuristic Mumbai skyline” really make someone the legal author of a highly complex, museum-quality piece of art? The user did not select the color contrast, the lighting, or the brush strokes—the algorithm did. Granting authorship for a simple text command dilutes the value of true human creativity.

The AI System Itself: Some radical tech scholars suggest recognizing the software itself as the author. Yet, under current Indian jurisprudence, an AI has no legal personality. It cannot hold property, it cannot sign licensing contracts, it cannot collect royalties, and it cannot be sued in a court of law for copyright infringement. You cannot grant legal rights to an inanimate mathematical code.

The Parent Company: The final view is that the technology corporation that owns the AI platform (like OpenAI or Google) should automatically hold the copyright. This approach completely corporate-izes authorship and creates a dangerous monopoly, allowing large tech giants to lock down millions of creative outputs without resolving the ethical question of true creativity.

Ultimately, our existing provisions fail because they are completely unequipped for AI-generated works. The law forces us to desperately search for a human owner, even when the human involvement in the final expression is practically non-existent.

V. Originality and Human Authorship Requirement

The second major legal hurdle that paralyzes our current system is the concept of “Originality” under Section 13. Under global intellectual property principles, copyright is not granted to every piece of text or image; it must possess originality. Historically, copyright law has always required an intellectual effort, a human spark, or a “modicum of creativity” originating directly from a human mind.

The core issue with modern generative AI is autonomous generation. When a human user sends a prompt into an AI engine, the human steps back completely, and the machine takes over the driver’s seat. The algorithm independently makes thousands of micro-creative decisions—choosing the sentence structures, selecting the vocabulary, blending the color palettes, and arranging the visual tones. This leads us to a massive, uncomfortable question: Can machine creativity ever satisfy the strict legal requirement of originality?

If our definition of originality strictly requires a human soul, human emotions, and human intellectual labor, then AI-generated works can never be legally original. This means they will be left completely unprotected, falling straight into the public domain where any corporation can copy and exploit them without permission.

But if we define originality purely by looking at the uniqueness of the final output, then AI easily satisfies the test because its outputs look fresh, creative, and completely distinct. This creates a deep, fundamental conflict between the traditional human-centric philosophy of copyright law and the mechanical reality of modern technology.

VI. Judicial Perspective on Copyright and Originality

Since the Indian legislature has remained completely silent on the rise of generative AI, we must look closely at how courts handle originality and authorship to guide our legal reasoning.

Indian Jurisprudence: The Modak Standard

In the landmark case of Eastern Book Company v. D.B. Modak (2008) 1 SCC 1, the Supreme Court of India fundamentally shifted our national standard of originality. Before this case, Indian courts followed the old English doctrine known as the “sweat of the brow,” which granted copyright protection to almost anything as long as the creator proved they had put in hard work, time, and money.

However, in the Modak ruling, the Supreme Court rejected this simplistic “sweat of the brow” doctrine and proudly adopted the “modicum of creativity” approach. The Court held that for a work to be original, it cannot be a product of mere manual labor or routine compilation. It must necessarily possess a minimal degree of creativity, and it must show the exercise of human skill, selective judgment, and creative choice.

Legal analysis: Applying the Modak standard to modern AI systems creates a massive legal paradox. An AI-generated painting or research article definitely possesses a visual or structural “modicum of creativity”—it looks highly sophisticated. But the reality is that this selective judgment and creativity belong entirely to the mathematical algorithm, not to the human user who typed a short prompt. Therefore, under current Indian case law, AI works sit in a dangerous gray zone. They are creative, but since that creativity does not flow from human skill and judgment, they fail our highest judicial test.

Foreign Jurisprudence: Monkeys and Human Sparks

To understand the global direction of intellectual property rights, we can look at two crucial foreign developments that heavily influence Indian legal thinking:

Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991): The US Supreme Court firmly established that copyright protection absolutely requires independent creation plus a modicum of a creative human spark. This was further solidified recently in the US case of Stephen Thaler v. Shira Perlmutter (2023), where a federal court explicitly rejected a copyright application for an AI-generated image, stating that human authorship is a non-negotiable, unshakeable requirement of the law.

Naruto v. Slater, 888 F.3d 418 (9th Cir. 2018): Popularly known as the “Monkey Selfie Case,” a crested macaque monkey took a series of beautiful photographs of itself using a photographer’s camera. The US Court of Appeals ruled that non-human entities (including animals and, by extension, autonomous software programs) do not have legal standing to claim copyright protection.

These landmark cases show that globally, courts are fiercely and stubbornly protecting the “human-only” boundary of copyright law, refusing to grant legal ownership to non-human creators.

VII. International Perspective on AI Copyright

Different jurisdictions across the world are tackling this technological crisis in various ways, and India can learn immensely from this diverse global landscape:

United States (US): The US takes a very strict, traditional approach. If a work does not have a human creator, it gets zero protection and falls into the public domain. They refuse to recognize computer-generated works without direct, continuous human control.

United Kingdom (UK): The UK is far more progressive and tech-friendly. Section 9(3) of the UK Copyright, Designs and Patents Act 1988 explicitly protects “computer-generated works” where there is no human author. It avoids the philosophical debate and simply grants the copyright to the person who made the arrangements necessary for the creation of the work, protecting it for a term of 50 years.

European Union (EU) & China: The EU focuses heavily on AI safety and transparency under its recent EU AI Act, forcing tech companies to disclose if copyrighted data was used to train their models. Meanwhile, Chinese internet courts have shown a pragmatic willingness to grant limited copyright to AI outputs if the human user can prove they exercised significant creative choice and arrangement in configuring the prompts over multiple rounds.

Legal Analysis: India’s framework is currently stuck in the rigid US model without having the clear statutory exceptions of the UK model. We are left trying to apply outdated 20th-century laws to fast-moving 21st-century realities, which leaves our domestic tech sector highly vulnerable.

VIII. Challenges in AI-Generated Works

The complete failure of our legislature to update our legal framework creates several real-world commercial and ethical crises:

Massive Ownership Confusion: Businesses and startup marketing agencies using AI tools to design commercial logos, write ad copies, or develop software codes do not know if they actually own their assets. This leads to massive commercial vulnerability, as competitors can copy their designs with zero legal consequences.

Lack of Legislative Guidance: Without clear statutory definitions, tech companies, contemporary artists, and transactional lawyers are completely left in the dark, forced to rely on unpredictable, expensive, and slow case-by-case court battles.

Ethical Issues & Copyright Infringement Risk: Generative AI tools do not create art out of thin air; they are trained by scraping billions of copyrighted human artworks, books, and articles off the internet without the consent, credit, or financial compensation of the original creators. This is mass copyright infringement happening in plain sight during the AI training phase.

The main goal of highlighting these challenges is to expose the massive structural weaknesses of our current legal setup. We are staring at a system that is completely unequipped for the digital future.

IX. Need for Legal Reform in India

India needs to act immediately. We cannot wait for our courts to solve a systemic problem that requires a legislative pen. The following concrete, structured reforms should be introduced by the Parliament:

New AI Copyright Guidelines: The Indian Copyright Office should issue clear interim circulars clarifying the exact percentage of human intervention and prompt-curation required to claim copyright over an AI-assisted work.

Separate AI Legislation: We should insert a specific provision into the Copyright Act—similar to the UK model under Section 9(3)—that explicitly recognizes “computer-generated works” and clarifies who the legal owner is, thereby providing market certainty.

Implementation of a “Human Involvement Test”: Introduce a clear legal standard to measure the “creative control” of the human. If a human uses AI merely as an assistant (like using a smart tool in Photoshop), full copyright should be granted. If the AI does all the heavy creative lifting autonomously, a completely different, lesser form of protection (a shorter term of 10-15 years) should apply to prevent tech monopolies from locking down public creativity.

Ownership Clarification: The law should explicitly state that the copyright of an AI-assisted work belongs to the human user who directed, curated, and shaped the creation, provided they can prove substantial creative choice in their input.

Conclusion

To summarize the entire legal landscape, Artificial Intelligence has permanently and aggressively broken the traditional boundaries of copyright law. The old, comfortable assumption that only human minds can generate original art, music, or literature is officially dead.

Our current Indian law is entirely outdated and full of statutory gaps regarding ownership and originality. My final opinion is firm and uncompromising: the Indian law must be updated immediately to survive the AI era. Technology is evolving exponentially, and if our laws remain static and frozen in the 1950s, our legal system will become completely irrelevant to the digital economy. We do not need to ban AI creativity or fear technological progress; we need to build a modern, sharp, and flexible legal framework—one that fiercely protects the emotional and intellectual labor of human artists while safely embracing and regulating technological innovation.

Reference(S): 

1. The Copyright Act, 1957, No. 14 of 1957 (India).  
2. Eastern Book Company v. D.B. Modak [/I], (2008) 1 SCC 1 (India).  
3. Feist Publications, Inc. v. Rural Telephone Service Co. [/I], 499 U.S. 340 (1991).  
4. Naruto v. Slater [/I], 888 F.3d 418 (9th Cir. 2018).  
5. Stephen Thaler v. Shira Perlmutter [/I], No. 22-v-01564 (D.D.C. 2023).  
6. Copyright, Designs and Patents Act 1988, c. 48, § 9(3) (United Kingdom).  
7.   Daniel J. Gervais, [I] The Machine As Author [/I], 105 [I] IOWA L. REV. [/I] 2053 (2020).  
8.   Pamela Samuelson, [I] Allocating Ownership Rights in Computer-Generated Works [/I], 47  PITT. L. REV. [/I] 1185 (1986).  
9. Parliamentary Standing Committee on Commerce, Review of the Intellectual Property Rights Regime in India, 161st Report (Presented on July 23, 2021).  
10. U.S. Copyright Office, Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence, 88 FR 16190 (Mar. 16, 2023).