AI-GENERATED WORK AND COPYRIGHT OWNERSHIP

Introduction

Artificial intelligence has unsettled one of copyright law’s most fundamental assumptions: that creativity is exclusively human. When systems like ChatGPT or MidJourney generate novels, images, or music with minimal human input, the law confronts a paradox[1]. Copyright insists on human authorship, yet the outputs of these systems often display originality and value comparable to human works. The question is not simply whether AI can be an “author,” but whether copyright law can remain credible if it ignores such creations.

The insistence on human authorship reflects a deeper normative choice: copyright is designed to reward human labor, moral rights, and personal expression. But this framework falters when confronted with autonomous machine outputs. Denying protection risks leaving valuable works in a legal vacuum, potentially discouraging investment in AI-driven creativity. Granting protection, however, raises equally troubling consequences: who should own the rights — the programmer, the user, or the machine itself? Each option exposes contradictions in existing doctrines of originality, ownership, and accountability.

Different jurisdictions illustrate this tension vividly. The United States Copyright Office has rejected AI-only works, clinging to human-centric definitions of authorship. The United Kingdom similarly emphasizes human creativity, while India’s silence leaves courts to improvise. These divergent approaches reveal not only doctrinal uncertainty but also a deeper policy struggle: should copyright evolve to accommodate technological realities, or should it defend its humanist foundations against dilution?

This article argues that the debate over AI-generated works is not about machines replacing humans, but about whether copyright can adapt without losing its legitimacy. By critically examining statutory provisions, case law, and policy debates across multiple jurisdictions, it will expose the gaps in current frameworks and evaluate possible reforms. Ultimately, the analysis will suggest that copyright must remain anchored in human creativity while recognizing hybrid models of authorship that reflect the collaborative nature of AI-assisted creation.

1. Legal Framework and Current Law

Copyright law is built on two interlocking concepts: authorship and originality. Both are traditionally defined in human terms. Statutes in India, the United States, and the European Union consistently tie authorship to “natural persons,”[2] embedding human identity into the very architecture of protection. This is not accidental — it reflects copyright’s normative purpose: rewarding human labor, safeguarding moral rights, and linking creativity to personality.

But this human-centric design is increasingly inadequate. AI systems now produce works that meet the traditional tests of novelty and distinctiveness, yet lack the human imprint the law demands.[3] The refusal to recognize such works exposes a contradiction: copyright claims to regulate creativity, but it excludes a growing category of creative outputs. This exclusion is not doctrinal necessity but a deliberate choice to preserve copyright’s humanist identity.

The analytical tension lies in how originality is defined. If originality is tethered to human personality — the “sweat of the brow” or “intellectual creation” doctrines — AI outputs cannot qualify. But if originality is understood functionally, as novelty or distinctiveness, then AI works arguably deserve protection. This reveals that copyright’s foundations are not neutral truths but philosophical judgments about what society values as “creativity.”

Another contradiction emerges in the treatment of computer-assisted works. When humans use software tools, their outputs are protected without question. Yet when AI systems generate works with minimal human input, protection is denied. The distinction is not technological but normative: law rewards human effort, not machine autonomy. This exposes a deeper fragility — copyright is less about regulating creativity than about defending a humanist conception of authorship.

Finally, the silence of statutes on AI-generated works is itself revealing. Legislatures have chosen not to address the issue, leaving courts and administrative bodies to improvise. This silence is not neutral; it reflects a reluctance to confront the philosophical challenge AI poses. By refusing to legislate, lawmakers preserve doctrinal purity but create uncertainty for businesses, investors, and creators navigating AI-driven industries.

2. Case Law Analysis

Courts and administrative bodies have confronted AI-generated works cautiously, often defending tradition rather than addressing technological realities. Their decisions reveal not only doctrinal rigidity but also deeper anxieties about copyright’s identity.

United States

The U.S. Copyright Office’s rejection of Zarya of the Dawn (an AI-generated comic book) illustrates this defensive stance. The Office declared that “human authorship is a bedrock requirement.”[4] This language is revealing: it elevates human authorship from a doctrinal rule to an existential anchor for copyright itself. The decision was less about the originality of the work — which was evident — and more about preserving copyright’s humanist foundation.

Analytically, this exposes a contradiction. Copyright law claims to regulate creativity, yet it excludes works that are undeniably creative simply because they lack human origin. The refusal is not about protecting creativity but about protecting copyright’s philosophical identity. In doing so, the law risks irrelevance in industries where AI outputs are already monetized.

United Kingdom

The UK’s Copyright, Designs and Patents Act takes a slightly different approach. It recognizes computer-generated works but still requires human authorship.[5] This compromise acknowledges technological realities but avoids the radical step of granting rights to machines.

Critically, this approach reflects a pragmatic balancing act. It allows protection for works where human input is indirect but still insists on a human anchor. Yet the compromise is unstable: if human authorship is minimal or tokenistic, does assigning rights truly reflect creativity? The UK’s model risks becoming a legal fiction — attributing authorship to humans who did not meaningfully create the work, simply to preserve doctrinal coherence.

India

India’s Copyright Act, 1957, is silent on AI-generated works. This silence is itself significant. It reflects legislative reluctance to confront the philosophical challenge AI poses. Courts may be forced to improvise, as they did in other contexts where statutes were silent, such as expanding privacy rights.

The analogy to Naruto v. Slater (the “monkey selfie” case in the U.S.) is instructive.[6] There, the court rejected non-human authorship even though originality was present. The reasoning was not about creativity but about preserving doctrinal coherence: authorship must remain human. If Indian courts face similar cases, they may adopt the same defensive posture, privileging doctrinal purity over technological adaptation.

Analytical Implications

These cases reveal a consistent pattern: courts cling to human authorship not because AI outputs lack originality, but because recognizing them would destabilize copyright’s normative foundations. The law is defending itself, not necessarily creativity.

This defensive posture creates contradictions. By denying protection, courts leave valuable works in a legal vacuum, undermining incentives for innovation. By insisting on human authorship, they risk creating legal fictions that attribute creativity to humans who did not meaningfully contribute. The result is doctrinal coherence at the expense of practical relevance.

Analytically, the case law demonstrates that the real struggle is not about whether AI can be creative, but about whether copyright can survive without its humanist anchor. Courts are choosing identity over adaptation, but this choice may render copyright increasingly disconnected from technological reality.

3. Critical Evaluation and Analysis

The refusal to recognize AI authorship creates a legal vacuum. Works of originality and value remain unprotected, leaving them vulnerable to exploitation. This undermines incentives for investment in AI-driven creativity and creates uncertainty for businesses.[7] Yet granting protection raises equally troubling consequences: corporations controlling AI systems could monopolize cultural production, sidelining human creators.

The accountability problem is central. Copyright assumes authors can bear responsibility for infringement, moral rights, and exploitation.[8] AI cannot. Assigning rights to machines is incoherent. Assigning rights to programmers or users risks over-rewarding them for creativity they did not produce.

This exposes a deeper contradiction: copyright law claims to reward human creativity, but in practice it often rewards capital investment. AI intensifies this contradiction. Protecting AI outputs may transform copyright from a humanist institution into a corporate monopoly system. The debate is therefore not about machines versus humans, but about whether copyright should remain a humanist institution or evolve into a tool of technological capitalism.

Analytically, this reveals that the real struggle is not doctrinal but philosophical: should copyright defend its humanist identity, or adapt to technological realities at the risk of becoming a corporate instrument?

4. Comparative Perspectives

Jurisdictions diverge sharply, reflecting different normative priorities:

• United States: Defends human authorship[9] rigidly, prioritizing doctrinal purity over technological adaptation. This reflects a fear of dilution: if machines can be authors, copyright loses its humanist anchor.
• European Union: Emphasizes human creativity but explores regulatory frameworks for AI, suggesting openness to hybrid models.[10] The EU’s approach reflects a willingness to adapt without abandoning humanist principles.
• India: Silence in statutes creates uncertainty, but judicial activism could fill the gap.[11] India’s courts have historically expanded rights creatively (e.g., privacy as a fundamental right), so they may adapt copyright similarly. This flexibility could allow India to craft a uniquely balanced approach.
• China: More pragmatic, sometimes recognizing AI outputs in practice[12], reflecting a policy of encouraging technological innovation. This approach prioritizes economic growth over doctrinal purity.

This divergence reveals a global struggle: should copyright evolve to recognize non-human creativity, or should it defend its humanist foundations? The comparative analysis suggests that no jurisdiction has found a stable solution. The debate is not doctrinal but philosophical: what values should copyright protect in the age of machine creativity?

Conclusion

The debate over AI-generated works and copyright ownership exposes a fundamental tension between law’s humanist foundations and technology’s disruptive realities. Current frameworks, anchored in the requirement of human authorship, exclude autonomous machine outputs. This exclusion is not merely doctrinal but normative: copyright law was designed to reward human labor, protect moral rights, and preserve the link between creativity and personality. Yet, as AI systems increasingly produce works of originality and commercial value, the refusal to recognize them risks undermining copyright’s credibility as a system of regulating creativity.

Case law and administrative decisions reveal a defensive posture. The United States Copyright Office’s rejection of AI-only works, the United Kingdom’s insistence on human authorship, and India’s statutory silence all reflect an unwillingness to confront the reality of machine creativity. These positions preserve doctrinal purity but fail to address the practical consequences: valuable works left unprotected, uncertainty for businesses, and the risk of exploitation without accountability. Conversely, granting copyright to AI outputs risks monopolization, shifting the system from a humanist institution to a corporate tool of technological capitalism.

The comparative analysis demonstrates that no jurisdiction has found a stable solution. The divergence is philosophical rather than doctrinal: should copyright evolve to recognize non-human creativity, or should it defend its humanist foundations against dilution? The answer lies in balance. Copyright should remain anchored in human authorship, but it must adapt to acknowledge hybrid models where human and AI collaborate. Legislatures should clarify authorship rules, perhaps through sui generis rights for autonomous AI outputs, while international bodies like WIPO should foster harmonization to prevent fragmentation.[13]

Ultimately, the challenge is not whether machines can be authors, but whether copyright can evolve without losing its legitimacy. By embracing reform that safeguards human creativity while recognizing technological realities, the law can remain both principled and relevant in the age of artificial intelligence.

BIBLIOGRAPHY

Primary Sources

• Copyright Act 1957 (India), available at Indian Kanoon.
• Copyright, Designs and Patents Act 1988 (UK).
• Case C-5/08 Infopaq International A/S v Danske Dagblades Forening [2009] ECR I-6569.
• Naruto v Slater 888 F 3d 418 (9th Cir 2018).
• US Copyright Office, Compendium of Copyright Office Practices (3rd edn, 2021).
• US Copyright Office, Policy Statement on AI-Generated Works (2023).

Secondary Sources

• Ding Y, ‘Artificial Intelligence and Copyright in China: Emerging Trends’ (2022) 12 Queen Mary J Int’l Prop L 45.
• European Parliament, Report on Intellectual Property Rights and Artificial Intelligence (2022).
• Lemley MA, ‘Intellectual Property and Artificial Intelligence’ (2020) 41 UC Davis L Rev 1.
• Samuelson P, ‘Copyright and Artificial Intelligence: The Challenge of Machine Authorship’ (2021) 68 J Copyright Soc’y USA 1.
• World Intellectual Property Organization, Issues Paper on Intellectual Property and Artificial Intelligence (2020).
• World Intellectual Property Organization, AI and Intellectual Property Policy Dialogue (2021).

[1] Pamela Samuelson, ‘Copyright and Artificial Intelligence: The Challenge of Machine Authorship’ (2021) 68 J Copyright Soc’y USA 1

[2] Copyright Act 1957 (India)

[3] Case C-5/08 Infopaq International A/S v Danske Dagblades Forening [2009] ECR I-6569

[4] US Copyright Office, Policy Statement on AI-Generated Works (2023)

[5] Copyright, Designs and Patents Act 1988, s 9(3)

[6] Naruto v Slater 888 F 3d 418 (9th Cir 2018).

[7] World Intellectual Property Organization, Issues Paper on Intellectual Property and Artificial Intelligence (2020)

[8] Mark A Lemley, ‘Intellectual Property and Artificial Intelligence’ (2020) 41 UC Davis L Rev 1

[9] US Copyright Office, Compendium of Copyright Office Practices (3rd edn, 2021) § 306.

[10] European Parliament, Report on Intellectual Property Rights and Artificial Intelligence (2022)

[11] Copyright Act 1957 (India)

[12] Yuanxi Ding, ‘Artificial Intelligence and Copyright in China: Emerging Trends’ (2022) 12 Queen Mary J Int’l Prop L 45

[13] World Intellectual Property Organization, AI and Intellectual Property Policy Dialogue (2021)