Can AI Own Its Creations?

Facebook
X
Facebook
WhatsApp

Can AI Own Its Creations? The U.S. Copyright Office Says No – But There’s More to the Story

By Eptain Editorial Team | April 2025

As artificial intelligence continues to transform how we create content—from art and code to music and marketing—a critical legal question has emerged: Can AI-generated work be copyrighted?

On April 2, 2025, the U.S. Copyright Office released Part 2 of its Report on Copyright and Artificial Intelligence, offering much-needed clarity. The conclusion: AI-generated work cannot be copyrighted unless there is significant human contribution.

Here’s a breakdown of what that means and why it matters for creators, developers, and innovators like us at Eptain.


Human Authorship is Essential

The report reinforces that U.S. copyright law only protects works created by humans. According to the Copyright Clause of the U.S. Constitution and decades of legal precedent, the term “author” refers to a person, not a machine.

So, if a piece of content is entirely generated by AI without meaningful human involvement, it does not qualify for copyright protection.


When AI-Generated Work May Still Be Protected

The Copyright Office outlines three specific scenarios where AI-assisted work may still qualify for protection—if the human element is clear:

  1. Assistive Use of AI
    When AI is used to support human creativity (like enhancing or editing), and the final product reflects human authorship, it may be copyrighted.
  2. Expressive Inputs and Human Influence
    If a human inputs original, copyrightable work into an AI system and the human expression is evident in the output, protection may apply—limited to the human-created parts.
  3. Creative Arrangement of AI Output
    If a human arranges, selects, or modifies AI-generated content in a creative way, the final result may be protected, as long as the changes meet the minimum threshold for originality.

Prompts Are Not Enough

Typing a detailed prompt into an AI system like ChatGPT or Midjourney doesn’t make the user the author of the result. The Office emphasized that the unpredictability of AI systems means users don’t have enough control over the outcome to claim authorship.

Even if the same prompt is used multiple times, the output can be entirely different each time. That lack of consistency makes it hard to establish a direct creative link between the prompt and the result.


Global Perspectives on Copyright for AI Works

The report also examined how other countries are handling the issue:

  • European Union: AI-generated content may be protected only if human input is significant in the creative process.
  • United Kingdom: Current laws allow for protection of computer-generated works even without a human author, but updates are under review.
  • Japan: Takes a case-by-case approach based on user involvement, input complexity, and editing.
  • China: Recognizes the person using the AI as the author of the work.

The global conversation is still evolving, and the U.S. is closely monitoring international developments.


No New Laws (For Now)

Despite rising calls for change, the Copyright Office believes existing copyright laws are flexible enough to handle emerging AI technologies. While some have argued for stronger protections to support innovation, the Office warns that doing so could come at the expense of human creators.

“If authors cannot make a living from their craft… society would be poorer.”


What This Means for Eptain and the Tech Community

At Eptain, we’re building AI-enhanced ecosystems and applications where content creation is key. Understanding the legal framework around copyright is critical—not just for compliance, but for innovation.

Here’s what we recommend:

  • Ensure your workflows include meaningful human authorship.
  • Don’t rely solely on AI for final content if you want copyright protection.
  • Stay informed as laws continue to evolve—especially if your business model involves generative AI.

Looking Ahead

Part 3 of the U.S. Copyright Office’s report is expected to cover:

  • The legal implications of training AI models on copyrighted material
  • Licensing frameworks for AI systems
  • Potential liability for developers and end users

We’ll continue to follow these developments closely and share our insights with the Eptain community.

Scroll to Top