Generative artificial intelligence has changed the way we write, but it has also unleashed an unexpected backlash: more and more authors, editors, and platforms are seeking to distinguish human text from machine-generated text. The phenomenon not only affects the publishing industry, but also extends to social media and subscription services.where authenticity has become an increasingly valuable asset.
In recent months, several cases have highlighted the tension between the efficiency of AI and the desire to preserve one's own voice. From the removal of a novel due to suspicions of artificial intelligence use to the adoption of detection tools by major platforms, The sector is seeking a balance that is not yet clear.>. Next, we review the main events that are marking this new stage.
Writers versus the machine: the rise of the anti-AI style
Novelist Laura Brooke Robson, author of 'Love Is an Algorithm', has explained that since the popularization of the great language models, she has modified her way of writing. Robson avoids generic metaphors, three-item lists, and long dashestraits he associates with AI-generated text. “I have this urge to make small mistakes, like winks to the reader, to show that there’s a human behind the words,” he stated. His testimony aligns with that of other authors who practice what has been called an “anti-AI” writing style, based on idiosyncrasy and deliberate imperfection.
On the other hand, the case of Mia Ballard's horror novel 'Shy Girl' illustrates the consequences of accusations of AI use. Hachette publishers withdrew the book from the market in March following online speculation.Although the author denied using artificial intelligence, Ballard asserted that she hired an editor and that the manuscript was analyzed with detection software, and that her priority now is for her writing to feel "real and authentic." This episode has alerted agents and publishers to the difficulty of identifying machine-generated text.
AI detection: tools and limitations
Faced with the growing presence of texts generated by artificial intelligence, tools designed to detect them have emerged. These systems analyze linguistic patterns and offer estimates about the possible automated origin of a textHowever, they don't provide definitive proof. Both academic studies and digital platforms have pointed out that these detectors can produce false positives, meaning they can flag text written by a human as AI-generated. This uncertainty has led many companies to use these systems as a guideline rather than as an automatic rejection criterion.
At the same time, readers and users increasingly value authenticity. Several analyses agree that people prefer to read content created by other human beingsAnd the mere fact of knowing that a text has been generated by a machine reduces interest. This demand for originality is pushing writers and platforms to find ways to ensure that content is genuine, although the path to achieving this remains fraught with technical and ethical challenges.
The combination of these factors paints a picture in which human writing seeks to reaffirm itself against automation.While writers modify their habits to avoid sounding robotic, platforms deploy detection tools and courts debate the legal limits of using copyrighted works to train models. Authenticity has become a scarce commodity and, paradoxically, writing versus artificial intelligence It is forcing humans to be more human in their writing style.