The U.S. Copyright Office refused to register the images generated by an AI tool for a graphic-novel project, but it approved copyright protection for the writer’s text and the way the pictures are arranged. The decision spotlights how the law draws the line between human-created work and output that comes straight from a machine, and it forces anyone who leans on AI for writing or illustration to rethink who actually owns the result.

Why the ruling matters

AI-driven text generators and image models have moved from hobbyist experiments to mainstream creative aids. Writers feed a prompt, tweak the output, and publish a finished piece in hours. Publishers, advertisers, and independent creators are betting on the speed and cost savings these tools promise. If the Copyright Office says an AI-generated image is not protectable, the creator cannot enforce exclusive rights or collect royalties, and anyone can copy or remix the image without permission. The same creator can still claim ownership over the words they wrote and the editorial choices they made in laying out the graphic novel.

The office’s policy is clear: a work is eligible for copyright when a human exercises “creative control” over the material. When a creator types a single prompt and lets the AI produce the final output, the office treats the result as lacking human authorship. In the recent case, the writer used Midjourney for a graphic novel. The Copyright Office protected her text and her arrangement of images, but refused to protect the images themselves.

The stakes for creators

  • Writers and illustrators – Document the extent of your input. The more you can show that you selected prompts, edited outputs, and arranged the final product, the stronger your claim to ownership.
  • Publishers – Contracts that assume full ownership of AI-generated content may need to be renegotiated or clarified to reflect that the underlying images are not protectable.
  • AI developers – Companies that provide text or image generators could face pressure to embed tools that record user edits, creating a trail of human contribution for copyright claims.
  • Legal practitioners – The decision adds a concrete example to a body of guidance that has been largely theoretical, giving lawyers a precedent to cite when advising clients.

What the office actually said

The ruling centered on the concept of “human authorship.” Copyright law protects original works of authorship that stem from a human mind. The office found that the writer’s role in shaping the story and ordering the panels satisfied that requirement for the narrative text and the overall layout. The images themselves were produced by the AI and fell outside the statutory definition of authorship.

Bottom line

For now, the Copyright Office draws a hard line: if you simply press “run” on a prompt and publish the output, you do not own it. If you steer the process—choosing ideas, ordering material, editing the result—you can claim ownership of the parts that reflect your creative decisions. As AI tools become more autonomous, that boundary will be tested again, and the legal framework will have to keep pace. Creators who want to protect their work should treat AI as a powerful instrument, not a co-author, and keep a clear record of the human choices that shape the final product.