Fan Fiction, Copyright, and the Challenges of AI

Once confined to niche communities, fan fiction has become a leading force in contemporary storytelling. Platforms like Archive of Our Own (AO3), with the viral reach of TikTok, have forced fan fiction to the forefront of both entertainment culture and legal battles. Fans create works out of love, but copyright law creates legal tension between authors.

What is Fan Fiction?

Meriam Webster defines fan fiction as “stories involving popular fictional characters that are written by fans and often posted on the Internet.”[1] But, fan fiction is much more than a story, it is the core of a community. It is inspiration and fantasy, bringing life to your favorite characters from books, movies, and even famous public figures.[2] Rebecca Tushnet captured the beauty of fan fiction in a simple analogy to Barbie.[3] In Tushnet’s example, a young girl is playing with her Barbies – creating adventures, elaborate scenarios, and improvising a dialogue.[4] Once the young girl puts her adventures onto paper, they become “fan” fiction.[5]

The concept of fan fiction became firmly cemented in 1966, when Jean Rhys created a story based on Charlotte Brontë’s Jane Eyre.[6] In it, Rhys reimagined the life of her favorite character’s wife.[7] That same year, Tom Stoppard developed a play for two side characters in Shakespeare’s Hamlet; and, fans of Star Trek created numerous works devoted to their favorite characters.[8]

In more recent times, increasing popularity has pushed fan fiction from the margins into the public eye. This change was largely fueled by advances in technology that have created new platforms for storytelling such as TikTok, Archive of Our Own (AO3), and Wattpad.[9] These platforms make fan fiction easier than ever to produce and distribute, transforming it into a cultural force.[10] Yet, with this new visibility comes legal scrutiny.

Intellectual Property Implications & Framework

To understand why fan fiction might conflict with the law, it is useful to outline the basics of copyright. The Copyright Act of 1976 allows literary works to receive copyright protection if they are “original works fixed in any tangible medium.”[11] Effectively, typed, written, published, and unpublished works qualify for copyright protection.[12] Copyright owners have special, exclusive rights to their work.[13] This includes the right to authorize the preparation of derivative works.[14] Fan fiction falls into the derivative works category, because it is work produced from or inspired by existing works.[15]

However, under the fair use doctrine, fan fiction writers are allowed to use copyright protected materials without permission from the copyright owner, but only if it meets certain criteria.[16] Fair use protection is not automatic, but rather applied on a case-by-case basis.[17] The most common types of cases involve legal actions based on extreme similarities between the derivative and original works.[18]

Enforcement of copyright protection depends heavily on an author’s preference.[19] For example, Anne Rice – author of The Vampire Chronicles – sent cease-and-desist letters to several websites, requesting that they remove all fan fiction related to her works.[20] Compare Rice’s actions with Neil Gaiman’s – the author of Coraline; he embraces fan fiction.[21] On Gaiman’s website, he published that fan fiction does not bother him, and he sees no harm in fan fiction given how many people enjoy it.[22] J.K Rowling’s approach to fan fiction lies somewhere in between Rice’s and Gaiman’s. The Harry Potter author encourages fan fiction, but subject to conditions – the fan fiction must remain online and non-commercial.[23] For example, in 2008, Rowling sued a fan over a derivative work; the fan had printed a Harry Potter encyclopedia without Rowling’s permission.[24]

Current Challenges: AI & Fan Fiction

With the emergence of artificial intelligence (AI), copyright law has grown even more complex. Traditional concerns surrounding plagiarism and unauthorized use have shifted towards questioning established ideas surrounding authorship, ownership, and the legality of training datasets.[25] Database training is the process of “feeding” or teaching information to AI programs, so that these AI platforms have sufficient knowledge and material to form responses, recognize patterns, and understand language.[26] When AI companies extract material from the internet without consent from original authors, this is known as “data scraping.”[27]

These issues came to the forefront in 2023, when AO3 writers discovered that OpenAI’s GPT-3 system was trained on their works, when the model produced the term “Omegaverse.”[28] Because “Omegaverse” is a term used almost exclusively within niche fandom communities, it rarely appears elsewhere online; writers recognized it as evidence that their stories had been used to train AI platforms.[29] Unfortunately, this would not be the last time AO3 writers experience their works being used without consent for the purposes of training AI platforms.[30] In April 2025, millions of works from AO3 writers were, again, scraped to train AI programs.[31] AO3 writers first learned about the scraping through a trending Reddit thread.[32] In response to the leak, a Tumblr user built a search engine for AO3 writers; there, they could input their usernames to see if their stories had been taken for AI development.[33]

The AO3 scraping scandals show just how blurry the lines have become between fandom creativity and legal, copyright issues surrounding AI.[34] The main concern is not only who scraped what information, but who may actually claim ownership rights to protect that information, especially when AI platforms start spitting out their own fan fiction. There are three possible contenders for the owner of AI-created fan fiction: the fan who prompted the AI model to create the fan fiction; the AI system itself; and the AI company running the platform.

Who Owns AI Generated Fan Fiction: The Prompter?

Does drafting a prompt for an AI program give a fan sufficient copyright protection? In an attempt to help answer this question, the United States Copyright Office published a guide in 2025, elaborating on what is protected when AI is involved.[35] According to the Copyright Office, the use of AI tools does not affect the availability of copyright protection, as long as AI is not substituted for human creativity.[36] Further, the Copyright Office provided that copyright protection is not available for purely AI-generated materials – including prompts.[37] These guidelines are only suggestions as to what can receive copyright protections; each case must be individually analyzed.[38] No clear boundaries are set, and the open-endedness of these guidelines highlights the deficiencies in our legal system.[39] For example, these guidelines still leave important questions unanswered: how much human creativity is needed; and, if prompts alone do not qualify for copyright protection, then who gets credit for the resulting work?

Who Owns AI Generated Fan Fiction: AI Program Itself?

Some might look to the AI system itself, treating it as the creative force behind the story. But here, the law has drawn a hard line: Machines cannot be authors.[40] This was made clear in Thaler v. Perlmutter, when an individual attempted to get copyright protection for an AI program.[41] Stephen Thaler created an AI program called “Creativity Machine”; it produced an artwork called “A Recent Entrance to Paradise.”[42] Thaler then submitted a copyright registration application to the U.S. Copyright Office, listing “Creativity Machine” as the sole author.[43] The court held that the Copyright Act requires human authorship in order for a work to be eligible for copyright protection.[44] Therefore, the Creativity Machine, as a non-human entity, could not be recognized as an author under the act.[45] But, the court did not answer whether Thaler himself could be considered an author for copyright protection of “A Recent Entrance to Paradise.”[46] One thing is clear: no matter how sophisticated a program is, a machine cannot legally “own” what it creates.[47]

Who Owns AI Generated Fan Fiction: AI Company?

That leaves a final possibility, the company that built and runs the AI program. However, this raises questions about ownership. For example, is the owner of the AI platform really the owner of stories created by the platform; and, are they still the owner if the stories were created using the work of other authors?[48] These questions highlight a tension at the heart of the debate; fan fiction authors are treated as infringers (based off the original authors’ works), while AI companies are free to exploit the works of these authors for training databases.[49] To complicate matters further, most platforms claim they do not own user outputs and disclaim responsibility if those outputs infringe on someone else’s work.[50] Our current legal framework is the culmination of these questions and tension: No ownership – not by the prompter, not by the AI itself, and not by the AI company.[51]

Conclusion

Fan fiction embodies creativity and community. It allows members of a fandom to embrace new stories and to create new plots for their favorite movies, books, and public figures.[52] But fan fiction has always existed in a legal gray zone; works are loved by fans, hated by authors, and tolerated or suppressed depending on who owns the copyright.[53] The rise of AI has only amplified this tension.[54] Works once confined to niche communities are now scraped into massive AI training datasets, producing AI-generated stories that belong to no one.[55]

As AI reshapes fan fiction culture,[56] the unresolved questions regarding ownership hang heavy. Should the law protect fans who provide detailed and creative prompts, or should it draw finer lines against machine-made content? Should companies have more stringent protocols to screen fan-created works from their databases?

It is clear that fan fiction now sits at the heart of a much larger debate over authorship and fairness in the digital age. Whether courts or policymakers choose to step in, fan fiction has become impossible to ignore, not just as entertainment, but as a paradigm for how we define or limit creativity itself in light of growing technology.

[1] Fan Fiction, Merriam-Webster, https://www.merriam-webster.com/dictionary/fan%20fiction (last visited September 30, 2025).

[2] An Introduction to Fanfiction, Leiden Univ. Ctr. For Digit. Humanities (Oct. 7, 2024), https://digmedia.lucdh.nl/2024/10/07/an-introduction-to-fanfiction/

[3] Rebecca Tushnet, Legal Fictions: Copyright, Fan Fiction, and A New Common Law, 17 Loy. L.A. Ent. L.J 651 (1997).

[4] Id.

[5] Id.

[6] Charlotte Brontë, Jane Eyre (1846).

[7] Anne Jamison, Introduction to Fic: Why Fanfiction Is Taking Over the World xi (2006).

[8] Id.

[9] Salome Bours, The Positive Impact of FanFiction: A Deep Dive Into Wattpad and AO3, The Lexington Line (Apr. 3, 2024), https://www.thelexingtonline.com/blog/2024/1/31/db-ao3s-impact-on-para-social-relationships.

[10] Id.

[11] 17 U.S.C. § 102.

[12] Madison Kennedy, The Role of Copyright Law in Fanfiction, St. Louis Volunteer and Accountants for the Arts (Aug. 7, 2024), https://vlaa.org/the-role-of-copyright-law-in-fanfiction/.

[13] 17 U.S.C § 106.

[14] Id.

[15] Kennedy, supra note 11.

[16] Id.

[17] Id.

[18] Conan Props. Int’l LLC v. Sanchez, No. 17-CV-162, 2018 WL 3869894 at *12 (E.D.N.Y Aug. 15, 2018).

[19] Kimberlee Weatherall, Author-Centered Copyright Enforcement?, 41 Colum. J.L. & Arts 545, 545-47 (2018).

[20] Katrina Monica C. Gaw, Restoring Balance to the Force (of Fandom): An IP Management Strategy for Walking the Fine Line between IP Protection and Fan Engagement, 62 Ateneo L.J. 1483, 1485-86 (2018).

[21] Meredith McCardle, Fan Fiction, Fandom, and Fanfare: What’s All the Fuss? Note, 9 B.U.J Sci. & Tech. L. 433, 450 (2003).

[22] Id.

[23] Grace Westcott, Friction Over Fan Fiction: Is this Burgeoning Art from Legal?, Literary Rev. of Can. (Aug. 2008), https://reviewcanada.ca/magazine/2008/07/friction-over-fan-fiction/.

[24]Warner Bros. Entm’t Inc. v. RDR Books, 575 F. Supp. 2D, 513 (S.D.N.Y 2008).

[25] Oversight of A.I.: Rules for Artificial Intelligence: Hearing Before the Subcomm. on Privacy, Tech., and the Law of the Comm. on the Judiciary, 118th Cong. 166 (2023).

[26] Eugenio Caterino, What is AI Training Data? Examples, Datasets and Providers, Datarade, https://datarade.ai/data-categories/ai-ml-training-data

[27] Natalia Demianenko, Introduction to AI Scraping: The Next Generation of Data, The Medium (April 21, 2024), https://medium.com/@n-demia/introduction-to-ai-scraping-the-next-generation-of-data-e23a8cdeba23.

[28] AO3 vs AI: Are Copyright Claims the Solution to Unauthorized Data Scraping of Fanfiction Sites?, The Univ. of B. C. (Dec. 22, 2023), https://iplaw.allard.ubc.ca/2023/12/22/ao3-vs-ai-are-copyright-claims-the-solution-to-unauthorized-data-scraping-of-fanfiction-sites/?utm_.

[29] Id.

[30] Decca Muldowney, Fanfiction Writers Battle AI, One Scrape at a Time, The Verge (Jun. 23, 2025, 5:30AM), https://www.theverge.com/ai-artificial-intelligence/688640/fanfiction-ai?utm_.

[31] Id.

[32] Id.

[33] Id.

[34] Id.

[35] Brent Lutes, Identifying the Economic Implications of Artificial Intelligence for Copyright Policy: Context and Direction for Economic Research, U.S. Copyright off. (Feb. 2025), https://www.copyright.gov/economic-research/economic-implications-of-ai/Identifying-the-Economic-Implications-of-Artificial-Intelligence-for-Copyright-Policy-FINAL.pdf

[36] Id.

[37] Id.

[38] Id.

[39] Joshua Gans, The US Copyright Office is Anti-Prompt, Joshua Gans’ Newsl. (Apr. 4, 2025), https://joshuagans.substack.com/p/the-us-copyright-office-is-anti-prompt.  

[40] Thaler v. Perlmutter, 130 F.4th 1039, 1045 (2025).

[41] Id. at 1043.

[42] Id.

[43] Id.

[44] Id. at 1045.

[45] Id.

[46] Id. at 1044.

[47] Id. at 1041.

[48] Sinbay Tan, AI Fanfare & Fanfiction: Do Fanfiction Writers Have Protections Against Artificial Intelligence, 90 Brooklyn L.R. 333, 334 (2024).

[49] Id. at 333-34.

[50] Peter Henderson & Mark A. Lemley, The Mirage of Artificial Intelligence Terms of Use Restrictions, 100 Ind. L.J. 1327, 1129-30 (2025).

[51] Copyright Ownership of Generative AI Outputs Varies Around the World, Cooley Alert (Jan. 29, 2024), https://www.cooley.com/news/insight/2024/2024-01-29-copyright-ownership-of-generative-ai-outputs-varies-around-the-world.

[52] An Introduction to Fanfiction, supra note 2.

[53] Weatherall, supra note 19 at 546.

[54] Sinbay, supra note 48.

[55] Muldowney, supra note 30.

[56] Sinbay, supra note 48.