---
title: "Writing in the Age of Artificial Intelligence"
description: "In early 2026, the publishing industry was rocked by a major scandal involving the alleged use of artificial intelligence (AI) in a novel by author Mia Ballard…"
url: https://www.independentpress.com/article/writing-in-the-age-of-artificial-intelligence
date: 2026-05-02
categories: ["Technology","Business"]
author: "Allie Pallotta"
---

# Writing in the Age of Artificial Intelligence

![Writing in the Age of Artificial Intelligence](https://images.ctfassets.net/ewtdlsoyixc1/1r0WNLq8S7S4494PAOpgKz/285915f182f4204c1e31c4c39e830dd7/Writing_in_the_Age_of_AI.png)

In early 2026, the publishing industry was rocked by a major scandal involving the alleged use of artificial intelligence (AI) in a novel by author Mia Ballard entitled _Shy Girl _that had been picked up by Big Five publishing imprint—Hachette Book Group. The story came to light after multiple social media posts and readers called out the novel for containing suspicious hallmarks of AI prose, which then appeared to be confirmed when AI detector Pangram scanned the text of the novel and found almost 78% to be AI-generated. The scandal led Hachette to cancel the book’s publication.

The presence of AI in creative spaces has become a controversial issue in recent years. Because large language models (LLMs) like ChatGPT, Claude, and Gemini can generate grammatically polished text in seconds, their rapid proliferation has led to an onslaught of AI-generated content on social media, blogs, and self-publishing platforms like Kindle Direct. Discourse over the use of AI in writing has been relatively confined to day-to-day cases, but the _Shy Girl _controversy marks one of the first instances where a major publishing house canceled the release of a book because of the suspected use of AI.

The scandal threatens to have a profound impact on the publishing industry and writing infrastructure. Currently, the U.S. Copyright Office maintains that pure generative output is ineligible for copyright protections; there needs to be clear indicators of human authorship or guidance for a work to be copyrightable. For publishers, most carefully assert that an author’s work should be original, but any concrete policies on regulating or banning the use of AI are left rather ambiguous. Further, publishers do not have an 100% foolproof way of determining whether a book was written with AI and thus have no strong enforcement mechanism to back up their policies. In addition, some publishers recognize that AI can be a tool for their employees, further complicating the stance they may be willing to take on the technology.

For writers, the discourse on the use of AI is more divisive. According to a survey of over 1,200 authors, roughly 45% of authors claim they use some form of AI assistance in their work. The most frequent uses are for research, marketing, outlining, and editing, with writers viewing it as a partner that can help make the writing process more efficient. That being said, around 74% of authors report that they do not disclose AI use. Some say that using AI for research, brainstorming, and editing should not necessitate disclosure, while others fear a negative reception among audiences if AI use is disclosed. Even still, some authors have chosen to use AI throughout the whole writing process. 

Author Coral Hart, a pseudonym, is a well-known novelist who has made six figures mass producing over 200 romance novels with the use of Antrophic’s AI assistant Claude. Hart has self-published under a total of 21 pseudonyms to avoid AI stigma. In an interview with the New York Times, Hart claimed that she was able to produce a book in forty-five minutes, asserting that this efficiency would win the race against writers who take months to complete their work. While Hart’s case is more a matter of quantity over quality, it emphasizes how the sheer volume of AI-generated books is flooding the market, which can create steeper competition for newer authors to gain recognition. Though there are writers like Hart who readily embrace AI, other authors have maintained staunch opposition. In an open letter to the Big Five publishers, many established authors, including big names such as Leigh Bardugo, Jodi Picoult, and Lev Grossman, asserted that they would not use AI in their books in any capacity. They claim that AI models have stolen the creativity of writers by training AI systems on copies of their books without their permission.

As the debate over AI in publishing continues, how are readers responding? For now, the responses from most readers are mixed. In one study by computer science professor Tuhin Chakrabarty, Chakrabarty found that readers sometimes preferred an AI-written passage that mimicked the style of an award-winning author over a passage actually written by the author themself. In a YouGov poll, however, research found that 61% of readers would feel somewhat or much less fulfilled if they learned a book had been AI-generated after finishing it. Outside of official research, many readers and social media users are becoming more sensitive to the telltale signs of AI writing, denouncing such writing as “slop.” While this allows readers to avoid engaging with AI-written work, in some cases, policing has been taken to the extreme in so-called “witch hunts,” where writers may be falsely accused of using AI. Nevertheless, readers’ general attitudes toward AI writing err on the side of caution and apprehension.

Unfortunately, for the publishing industry, the rapid evolution of AI models will make it increasingly difficult to recognize the use of AI in writing. At the very least, authors should be required to divulge the fact that they employ AI in their writing. Readers have the right to know. The use of AI should be fully disclosed by authors so that readers can decide if they want to engage with work that is AI assisted. After all, you would not use meat in a dish and knowingly serve it to a vegetarian without disclosing the ingredients. While AI-generated images and videos are increasingly identifiable via watermarks, there is currently no similar watermark for AI-generated writing. Most AI detectors are said to be unreliable, in addition to the fact that they use AI themselves to determine usage, which may turn people off from using them. Some authors choose to include a disclaimer at the beginning or end of their work stating whether AI was used and how, but that often relies on the honor system and word of the writer. A more impartial solution could be to embed and incorporate a symbol like a trademark “™” in generative outputs to clearly indicate AI usage.

Whether people think AI in writing is useful or ethical is still being decided by institutions and the court of public opinion. In the meantime, writing continues to be a space where people can express themselves and create and share new ideas with one another. Writing helps people build connections by seeing how different people imagine and interact with the world, and if that ability to be authentic and trust one another is lost or replaced, then we risk losing the soul and joy of the written word.