Executive Summary
The rapid adoption of generative artificial intelligence (AI) has raised fundamental questions about copyright ownership, authorship, and enforceability. Individuals and institutions increasingly rely on AI systems—such as large language models and image synthesis tools—to draft text, generate images, and assist with research. Yet, copyright law in most jurisdictions remains grounded in the concept of human authorship.
This whitepaper examines who owns copyright in works created using generative AI. It clarifies common misconceptions, analyzes prevailing legal doctrines across key jurisdictions, and distinguishes between AI-generated content and human-authored works created with AI assistance. The analysis demonstrates that copyright ownership hinges not on the use of AI per se, but on the presence of meaningful human creative control.
Key conclusions:
- Generative AI systems cannot own copyright.
- Purely AI-generated content is generally not protected by copyright.
- Human-authored works created with AI assistance are typically copyrightable, with ownership vesting in the human author or commissioning institution.
- Transparency and documentation of human involvement are increasingly critical for enforceability and credibility.
The paper concludes with practical guidance for creators, organizations, and policymakers navigating AI-assisted creative and research workflows.
Introduction: Copyright Meets Generative AI
Generative AI tools are now embedded in creative, academic, and commercial workflows. Text generation, image synthesis, and automated code drafting have moved from experimental to routine. This technological shift exposes a tension between existing copyright frameworks and new modes of creation.
Copyright law was designed to regulate human authorship, presuming an identifiable individual who exercises creative judgment. As AI systems produce increasingly sophisticated outputs, determining whether a work is protected—and who owns it—has become a pressing legal and policy question.
This paper addresses that question by focusing on ownership rather than infringement, originality, or training data. Its central inquiry is simple but consequential: when generative AI is used in the creative process, who owns the resulting work?
Foundational Principle: Human Authorship
Across most legal systems, copyright protection rests on human authorship. While statutory language varies, the underlying concept is consistent: copyright subsists in original works that reflect human intellectual creation.
Generative AI systems do not possess legal personality, intent, or creative agency. They cannot hold rights, bear duties, or be recognized as authors. Consequently, disputes over AI-generated works hinge on whether a human exercised sufficient creative control over the output.
This principle is the analytical anchor for modern approaches to AI-assisted works.
Pure AI-Generated Content and the Absence of Copyright
Where content is generated autonomously by an AI system without meaningful human input, most jurisdictions deny copyright protection. There is simply no human author to whom rights can attach.
Key implications:
- No exclusive rights arise in AI-only content.
- The content may be freely reused by others.
- Claims of ownership or exclusivity are difficult to sustain.
This reflects a deliberate policy choice: copyright rewards human creativity, not automated production.
Human-Led, AI-Assisted Works
Human-led creation supported by AI tools is the more common and legally significant scenario. Here, AI functions as an assistive technology rather than an autonomous author.
Examples:
- Drafting text based on human-defined prompts and outlines
- Editing, summarizing, or reorganizing human-authored material
- Generating preliminary drafts that are substantively revised by a human
Copyright protects the human contribution—the selection, arrangement, expression, analysis, and judgment reflected in the final work. Ownership typically vests in the individual creator or, where applicable, the employer or commissioning party under standard doctrines.
Case Study: AI-Assisted Whitepaper Drafting
A researcher uses a large language model to draft an initial outline of a whitepaper. The researcher then:
- Revises sections to incorporate original analysis.
- Reorganizes content to support a novel argument.
- Adds citations, commentary, and conclusions.
Outcome: Copyright subsists in the researcher’s contributions, not in the AI-generated draft alone. Ownership may belong to the researcher or their institution if the work is commissioned.
Comparative Jurisdictional Approaches
United States: Copyright protects only works of human authorship. The U.S. Copyright Office explicitly states that AI-generated content without human input is not registrable. Human-authored portions of AI-assisted works may qualify if they meet originality standards.
European Union: EU copyright law protects works reflecting the author’s own intellectual creation. Purely AI-generated works are excluded, but AI-assisted works are protected where human creative choices are evident.
United Kingdom: The UK recognizes limited computer-generated works, assigning authorship to the person who made the necessary arrangements. The scope of this protection in complex AI contexts remains largely untested.
Other Jurisdictions: Most common-law and civil-law systems follow the human authorship model. Statutory language varies, but outcomes are similar in practice.
Who Does Not Own Copyright
Contrary to some misconceptions:
- AI developers do not automatically own rights to user-generated content.
- AI systems themselves cannot hold copyright.
- Ownership is not transferred merely by using an AI tool.
Instead, ownership depends on contractual arrangements and the nature of human involvement in the creative process.
Practical Implications for Creators and Institutions
To strengthen copyright claims and maintain credibility:
- Maintain human editorial control over AI-assisted outputs.
- Avoid publishing AI-generated content verbatim when exclusivity is desired.
- Document creative decision-making, revisions, and AI interactions.
- Disclose AI assistance where appropriate.
Policy Considerations and Open Questions
As generative AI becomes more capable, several policy questions merit attention:
- Should new, sui generis protections for AI-assisted works be considered?
- How can innovation incentives be balanced with public access?
- Should disclosure obligations be standardized?
This paper emphasizes clarity and restraint, without advocating immediate legal reform.
Conclusion
Copyright ownership in the age of generative AI is less radical than it may appear. Existing legal principles remain largely adequate when properly applied. The decisive factor is not the presence of AI, but the role of human creativity.
Where humans lead and AI assists, copyright endures. Where automation replaces authorship entirely, copyright recedes. Understanding and respecting this boundary is essential for creators, institutions, and policymakers.
(The views and opinions expressed in this article are solely those of the author and do not necessarily reflect the official policy or position of any organization or entity.)
Disclaimer: This article is for general informational purposes only and does not constitute legal, technological, or professional advice. Laws and regulations vary by jurisdiction; readers should consult a qualified professional for advice specific to their situation.
While every effort has been made to ensure the accuracy of the information provided, readers should be aware that information is inherently dynamic. Laws, regulations, technology, etc., may change over time, and the author assumes no responsibility for errors, omissions, or outcomes resulting from the use of this information.
Links to external websites are provided for convenience and do not constitute endorsement.
