It is something to be thought through carefully, because technology is slipping out of our hands and we could end up being its toys. After threats of total destruction and apocalyptic statements about the so‑called Artificial Intelligence, other, smaller problems have begun to emerge, but none the less pressing for that. And the publishing industry has now been touched. A few days ago the Mexican imprint Almadía announced that it would suspend all activities related to the presentation of the book Donde el agua trabaja, by Peruvian writer David Silva Rojas (1981), winner of the I Premio de Novela Breve Almadía-Ventosa Arrufat, the result of which was announced last July. In a statement, the editors sought to address a news item from a Spanish newspaper in which it was disclosed that 99% of the novel’s text would have been generated with artificial intelligence (AI), according to Pangram, a detector (I suppose also intelligently artificial) of texts produced by machines. For the editors, publishing a book, they said, implies a relationship of trust among the writer, the editor, and the reader, trust that, they assure, “is the base on which Almadía is built” and, therefore, they have wished with that decision to “safeguard the commitments that make it possible,” since they claimed to be unsettled by the possible use of generative artificial intelligence in the writing of the winning novel, a situation that, they noted, opened questions “that must be analyzed seriously.” Very well. Almadía recalled that, since the call for entries was opened, the prize’s rules and protocol already established the prohibition of using artificial intelligence to write the participating works, so when naming the prize winner, they had him declare in the contract that he had not resorted to these tools, something that he also assured them personally and, they claim, there were no elements to doubt his word. On the other hand, there were several readers of the novel, such as the Jury itself (which included authors of proven literary discernment such as Cristina Rivera Garza, Selva Armada and Mónica Ojeda), and no one detected signs that would lead them to think that generative AI had been used. So the assessment of the novel focused on the work as a whole, its narrative voice, its construction, and the meanings it proposes, aspects they consider essential for approaching literature. Almadía recognizes that the rapid evolution of AI tools poses new challenges for the publishing sector and obliges us to reflect on the limits of their use, as well as on the trust placed in detectors, which are also AI, albeit in this case not generative but audit, watchful, supervisory, supervisory, evaluative, verifiable, examinative and tattling, if you’ll allow me. In short, Almadía, with good judgment, state that these tools can provide information, but their results must be interpreted and cross‑checked, especially when the career of an author and the readers’ trust are at stake, so in the face of such a scenario the publishing house has deemed it necessary to carry out a detailed review of the matter before reaching a final conclusion about the case, but while they carry it out, they have suspended all activities related to presenting the book. For his part, the then‑author of Donde el agua trabaja, a work that addresses labor exploitation and contemporary forms of vulnerability, with silence as a compositional element, according to the description released by Almadía when announcing the award, has sought to respond to this matter and asserts and reiterates that he is the author of the novel in question, and on a social network he stated that he conceived, wrote, and revised it without using generative AI tools to create it. “The novel,” he said, “sprang from an investigation into the living and working conditions of people who come to Ica (a region in southwestern Peru) to work in the agricultural sector. While transcribing the interviews, I began to write small fragments in the first person, gathering in an imagined voice experiences and sensations that those testimonies had left me. At first I did not intend to write a novel.” That decision, he notes, came later, when he decided to turn those fragments into a novel and prepared a notebook in which he laid out how the story would begin, what would be the main conflict and toward what ending he wanted to guide it. “I also noted some guidelines that would serve me during the writing, for example, always maintaining the protagonist’s perspective, using a restrained voice and working the rhythm through repetitions and silences.” That framework, he clarified, was the skeleton of the novel, not a detailed plan of all its chapters, but as he wrote, he was incorporating scenes, characters and situations that were not initially foreseen, but that allowed him to deepen the story without straying from its main direction. Since he did not need to register every new idea in the notebook, he continued to revise the structure and language of the manuscript in different Word versions, machine‑type versions that, he asserts, he preserves, as well as a notebook of notes and materials related to the research that gave rise to the novel, which is why he has been gathering and organizing that documentation “to hand it to the Almadía publishing house and collaborate with the announced review”. Finally, Silva Rojas has offered a reflection on the so‑called artificial intelligence, and maintains that it raises important questions for literature and for publishers. “However,” he points out, “I consider that the results of a detection tool, no matter how high its percentages, should not be taken as definitive proof of how a work was written without examining other elements.” And he closes by asking that the review be carried out with care and that its conclusions be supported by facts. “For now, my priority is to collaborate with the publisher and provide the documents that allow examining my writing process,” he concludes. This case is very important for several questions: will we blindly trust computation, information technology, and algorithms? who is smarter: a writer, an editor, a reader or a machine and its software? are we at the thresholds of a new era where we must face that there exist new tools, as engineers and creators did in other times? is a photograph of Venice more false than a Tintoretto landscape?