Mexican publishing house Almadía has announced its decision to uphold the outcome of the I International Short Novel Prize Almadía-Ventosa Arrufat, which was awarded to the novel Donde el agua trabaja. This determination followed a thorough evaluation of the work by Peruvian author David Silva Rojas, who faced public allegations of having utilised artificial intelligence (AI) to produce 99% of the manuscript.
In a statement released by the publisher, the Ventosa Arrufat Society, and the panel of judges—which includes Cristina Rivera Garza, Selva Almada, and Mónica Ojeda—it was emphasised that, “This award embodies the effort and trust of many individuals. In respect to the author, the judges, the Ventosa Arrufat Society, and the readers, Editorial Almadía has decided to conduct a review of Donde el agua trabaja. Upon its conclusion, the publisher reaffirms its confidence in David Silva Rojas and maintains the ruling of the I International Short Novel Prize Almadía-Ventosa Arrufat.”
A Notable Achievement Amid Controversy
The novel emerged victorious from a pool of over 2,200 manuscripts submitted from 22 countries. However, last month, an investigation published by a leading Spanish newspaper reported that Pangram, a tool designed to detect AI-generated texts, attributed nearly complete authorship of the novel to this technology. This revelation sparked a heated debate on social media, with some users sharing excerpts they believed exhibited characteristics typical of AI-generated writing, while others questioned the validity of a programme undermining an author’s reputation without substantial evidence.
In light of the controversy, Almadía chose to suspend the book’s promotional events and announced an internal investigation. Silva Rojas has categorically denied the use of AI in his writing process and expressed his willingness to provide materials related to his creative journey.
Thorough Examination and Ethical Considerations
In its statement on social media, the publisher clarified that the review was conducted using materials provided by the author. “Almadía based its judgement on human criteria, considering the entirety of the examined information, its connection to the work, and various discussions held with technologists, lawyers, and linguistic experts during the process,” it underscored.
Almadía acknowledged the journalistic interest surrounding the doubts cast on an award-winning novel, asserting that such concerns should be addressed with the utmost seriousness. However, it stressed that “the results of a detection tool alone are insufficient to conclusively determine how a work was created.” Furthermore, it noted that the aforementioned investigation “did not always sufficiently distinguish between suspicion and proven fact, which had repercussions for the author, the judges, and the publisher.” The publisher reminded that media outlets bear ethical responsibility for the consequences of their publications and suggested that the Spanish newspaper should issue an apology as a “gesture of respect and consideration.”
Concerns About AI’s Impact on Literary Creation
Additionally, the publisher raised questions about the implications of granting authority to a foreign entity, whose algorithms were primarily trained on English-language data, to validate works written in Spanish. “We are concerned that such authority may ultimately influence the act of writing itself. George Orwell’s character Syme remarked in 1984: ‘Each year there will be fewer words and the range of consciousness will become ever smaller.’ What happens, then, when someone avoids a word, a rhythm, or a technique out of fear of being judged by a machine?” the publisher stated.
In conclusion, Almadía reiterated that the use of AI as a detection tool has sparked a significant debate within the publishing industry. This discussion continues, and experts predict it may escalate due to the advanced capabilities of recent algorithms. “For our part, we prefer to continue placing our trust in human judgement,” the publisher concluded.
The case of Almadía highlights the potential repercussions that the irresponsible and uncritical use of AI as a verification mechanism can have on the reputations of authors, publishers, and seasoned voices in the literary field. This risk extends to various other creative sectors, where it is becoming increasingly challenging to ascertain the origin of works, which ironically are employed by companies to train their detection and content generation algorithms.