Who Owns an AI-Generated Story?

The Legal Position in 2026

The law on AI-generated content ownership has become clearer over the past two years, though it remains an evolving area. The foundational principle in the United States is that copyright requires human authorship. This is not a new rule created for AI. It has been the standard for over a century, applied consistently to cases involving animals, nature, and machines. What is new is its application to the output of large language models.

The key case is Thaler v. Perlmutter, in which Stephen Thaler sought to register copyright for works created by an AI system called DABUS. The district court ruled against Thaler, and the Supreme Court declined to hear the appeal in March 2026. This effectively confirmed that AI cannot be listed as an author on a copyright registration, and works produced entirely by AI without meaningful human creative input are not eligible for copyright protection in the United States.

The U.S. Copyright Office has issued guidance reinforcing this position. In its February 2023 registration guidance and subsequent policy statements, the Office stated that works generated by AI in response to a prompt are not copyrightable, because the human's contribution (the prompt) does not constitute the kind of creative authorship that copyright law protects. The person who writes a prompt is giving instructions, not creating the expressive content.

When Human Involvement Creates Copyright

The situation changes when a human makes substantial creative contributions to the final work. The Copyright Office has indicated that works containing both AI-generated elements and elements of human authorship may receive protection for the human-authored portions. This is the critical distinction for writers using AI story generators.

If you type a one-line prompt into an AI tool and publish the raw output without modification, that text likely has no copyright protection. Anyone could copy it without legal consequence. But if you use AI to generate a first draft and then substantially edit, revise, restructure, and rewrite the text, the resulting work may qualify for copyright protection based on your creative contributions.

The Copyright Office has not drawn a precise line defining how much human editing is "enough" to establish authorship. The standard is whether the human has exercised sufficient creative judgment in selecting, arranging, and modifying the material. Fixing a few typos probably does not meet this threshold. Rewriting dialogue, restructuring scenes, adding original passages, developing characters beyond what the AI provided, and making substantive creative decisions throughout the text probably does.

The practical guidance for writers is straightforward: the more of your own creative work you put into the final product, the stronger your ownership claim becomes. Writers who use AI as a brainstorming and drafting tool but invest significant effort in editing and revision are in a much stronger legal position than writers who publish AI output with minimal changes.

Practical Recommendations for Writers

The legal landscape is still settling, but writers can take practical steps to protect themselves and their work today.

First, treat AI output as raw material rather than finished product. The more you edit, revise, and add your own creative contributions, the stronger your ownership position. This is good advice for both legal protection and quality, since AI output almost always improves with human editing.

Second, keep records of your creative process. Save your prompts, drafts, editing notes, and revision history. If you ever need to demonstrate that you made substantial creative contributions to a work, documentation of your editing process supports that claim.

Third, read the terms of service of the AI tools you use. Most major platforms assign output rights to users, but some platforms retain certain rights, such as the ability to use your input and output for model training. If you are working on commercially valuable fiction, understand what rights you are granting the platform.

Fourth, be transparent with publishers and platforms. Concealing AI use when a publisher or platform requires disclosure can result in removal of your work, loss of revenue, and reputational damage. Honesty is both the ethical and the practical choice.

Finally, consider the copyright status of your work as a spectrum rather than a binary. A story where AI contributed 95% of the final text has a much weaker copyright position than a story where AI contributed a rough first draft that the writer then substantially rewrote. Aim for the latter if you want meaningful legal protection.