Readers Are Using Chatbots to Write Their Own Fiction, and Publishers Are Watching

A segment of readers has started using AI chatbots to produce fiction tailored to their own tastes, according to reporting from Wired, which profiled users who generate custom stories ranging from romance to niche fantasy. The practice marks a shift in how AI is entering publishing: rather than authors using the technology, it is readers writing for themselves. The trend matters because it reshapes who gets to be a storyteller and what demand for professionally published books might look like down the line.

The publishing industry has spent much of the past two years debating how writers use generative AI, with lawsuits, disclosure policies and contract clauses multiplying across the sector. Readers experimenting with chatbots sit outside that fight. They are not selling books or submitting manuscripts; they are prompting models to produce serialized chapters, personal fan fiction and stories built around very specific premises. Wired's headline example, a user devoted to a story about a large rat spouse, illustrates how far the subject matter can drift from what a commercial publisher would ever acquire.

The mechanics are straightforward. A reader supplies a premise, characters and tone, then iterates with the model over multiple sessions, refining plot turns and dialogue the way a writer might revise a draft. Chatbots from OpenAI, Anthropic, Google and a cluster of smaller apps can all handle this work, and some users treat the output as a private hobby rather than literature. That keeps the activity largely invisible to publishers, agents and the copyright questions that dominate industry conferences.

For traditional publishing, the implications are mixed. On one hand, readers who spend hours generating their own stories are spending hours not buying books, which could chip away at a market already fighting for attention against streaming, games and social video. On the other, the same readers are demonstrating appetite for long-form narrative and for genres that mainstream houses often underserve. Some in the industry may see a signal about what audiences want, even if the delivery method is unfamiliar.

The legal and ethical picture remains unsettled. AI models are trained on copyrighted text, and several authors have sued developers over exactly that, so reader-generated fiction sits on contested ground even when no money changes hands. Publishers have also begun adding AI clauses to contracts and experimenting with their own AI-assisted titles, which makes the reader side of the equation harder to dismiss as a fringe curiosity.

What happens next depends on whether this stays a personal pastime or becomes something more organized. If platforms add sharing features, subscription tiers or tools that let users publish and trade their generated stories, the line between reader and author could blur further. The open question for the industry is whether it treats these users as a threat to book sales or as evidence of unmet demand for the kinds of stories big publishers rarely take a chance on.