Amazon acquires and mass destroys books in Las Vegas for Artificial Intelligence training

Users visiting online e-book stores such as Amazon are encountering a large number of works created or written using artificial intelligence (AI). These works often feature standard covers and template plots. Today, authors use AI-based systems that promise a ready-made e-book in less than 30 minutes, allowing them to release hundreds of titles annually under numerous pseudonyms.

This influx of new publications is impacting the publishing industry. Readers complain about the low quality of 'robotic content,' while human authors are losing visibility in search engines and seeing a decline in their income. If the advent of the e-book previously simplified distribution, AI has now significantly lowered production costs, but the market has become oversaturated. The question arises as to how to find clarity amidst so much noise.

Maurício Pinheiro, a professor of technology and arts from Sesc Piracicaba, explains to the publication Olhar Digital that AI has changed the scale of independent authorship. Now, one person can conduct research, structure, write, edit, and format a work.

This allows the author to operate almost like a micro-publishing house, testing various covers, descriptions, and niches at a speed impossible under the previous model.

This simplicity has transformed the logic of the digital market. Previously, producers aimed to create bestsellers, but the focus has now shifted to overall sales volume. Since the cost of text generation has dropped sharply, publishing dozens of titles consecutively has become a profitable business, even if each book sells little or the income depends on pay-per-read in subscription services.

Industry data confirms the impact of this industrial production tactic. For instance, a study by the National Bureau of Economic Research (NBER) shows that the number of e-books released on Amazon monthly has nearly tripled between 2022 and the end of 2025.

Another study conducted by the State University of New York found that the number of books with registered sales increased 19.2 times during this period, while total market revenue only grew by 8.9 times. This means that far more works are competing for readers' attention, leading to a decrease in average income per book.

According to the professor, who is also a software analyst, the reduction in production barriers has eliminated traditional filters that screened books before they reached the public. The specialist notes: 'For a long time, publishing was complex, and this complexity served as a filter before the book reached the audience. Now the filter happens later.'

The virtual stores themselves are now struggling to control this flow of synthetic content. For example, on the Rakuten Kobo self-publishing platform, 45% of uploaded e-books were rejected in one year, with 85% of these rejections being due to the books being AI-generated. Despite blocks and analysis, automated systems continue to bypass detection and add titles to stores daily.

Beyond market consequences, mass production of AI-generated e-books faces a legal hurdle in Brazil. Lawyer Marcelo Gargano, a digital law specialist at ABE Advogados, explains to Olhar Digital that Brazilian copyright law focuses on a natural person and requires human creativity to protect a work.

Gargano asserts: 'An e-book entirely generated by AI without significant human creative input does not have a natural person who could be recognized as the author of the creation, and without an author, there is no copyright protection.'

This implies that these books can be freely copied by third parties and also carry other risks for those who publish them, as there is a danger that the tool might reproduce copyrighted excerpts from other authors.

Furthermore, there is a discussion about the reader's right to know the origin of the purchased product. However, the country does not yet require platforms to indicate the use of technology on covers. The lawyer points out: 'It can be assumed that the consumer has the right to adequate and clear information about various products and services in accordance with the CDC, and therefore, the mentioned authors and platforms that publish them should inform consumers that this work was created using AI.'

Amazon serves as an example. On this marketplace, the books Tumbleweeds & Tequila: A Forced Proximity Romance (Shots of Love Book 2) by Sydney Marsh and Fighting Love’s Flame (Alaska Rugged Heart Series Book 3) by Dana Winston carry the following notice: 'This story was produced using AI tools, directed by the author' (free translation). However, it should be noted that this warning is hidden at the bottom of the description, behind the 'read more' button.

This legal vulnerability is accompanied by a change in the very value of the book, beyond its monetary equivalent. Since algorithms are capable of generating stories in large volumes, the ability to create acceptable text is no longer a distinguishing feature. AI is very competent at creating what seems plausible, familiar, and good enough. But there is a curious consequence: in an ocean of 'good enough' things, 'good enough' ceases to be a rare quality.

In this scenario of abundance, the value of the independent market shifts towards what cannot be automated. Factors such as personal style, repertoire, life experience, and the certainty that there is a real author behind the pages gain more importance for the reader's choice. The Sesc Piracicaba professor analyzes: 'If the text becomes practically infinite, then what cannot be produced infinitely begins to be valued higher: trust, reputation, perspective, repertoire, lived experience, taste, and recognizable relationship between the work and someone who is willing to take responsibility for it.'

Ultimately, self-publishing has turned into a battle against being overshadowed in an infinite catalog. Pinheiro concludes: 'I do not think the main fight will simply be between the human author and the artificial author. The fight will be between signal and noise.'

For the reader, the task of choosing the next book to immerse oneself in becomes more difficult. For the human writer, creating something outstanding now requires a 'secret ingredient' that a machine cannot replicate.

Opposition to artificial intelligence (AI) is intensifying in both America and Europe. A recent incident occurred in the US capital, Washington, where demonstrators infiltrated the OpenAI office to protest the rapid development of AI. According to a report by Futureism, demonstrations took place in the lobby of the OpenAI office in Washington, lasting about two hours, resulting in the arrest of 13 students.

This protest goes beyond a typical demonstration outside one company's office. Recently, there has been growing dissatisfaction with various aspects of AI in the US and Europe. Some fear job losses, others are concerned about the use of their creative works, and a third group opposes the massive data centers being built for AI due to their high consumption of electricity and water.

Previously, in March 2026, protesters marched past the headquarters of OpenAI and Elon Musk's xAI in San Francisco. According to Reuters, they demanded a halt to AI development.

The demonstration in Washington was triggered by concerns over the rapid impact and potential damage posed by AI. Protesters argue that the pace of creating new AI technologies is outpacing the development of corresponding regulations and safety measures by governments. This is why various groups opposing AI in the US are now directing their demands against the offices of companies, government institutions, and data center projects.

In March, protests also took place in San Francisco near the facilities of OpenAI, Anthropic, and Google DeepMind. In July, people associated with StopAI marched against these companies in San Francisco, questioning the speed of AI development.

The most significant reason for the backlash against AI is the threat of job displacement. A study conducted at King's College London in 2026 showed that about seven out of ten people are concerned about the economic consequences of job loss due to AI. Fifty-seven percent of respondents believe that AI could lead to mass unemployment, and about twenty-two percent think that AI could destroy jobs so quickly that it would cause social instability.

Fear is particularly high among youth and junior professionals. The study noted that six out of ten people believe AI could seriously affect entry-level office positions. This fear is not limited to employees. Governments in Europe are also discussing how to preserve people's jobs in the age of AI and how to provide employees with new skills.

The second major front of protest involves artists and content creators. AI models require huge amounts of data for training, which can include books, photographs, music, articles, and other internet content. Artists and writers ask: if their work is used to train AI models, why are they not given permission or paid?

In Europe, this issue has become so serious that the European Parliament has emphasized the need for new norms regarding AI and copyright. The Parliament's report states that existing rules are insufficient for using copyrighted works in AI training, and calls for greater transparency from AI companies and better protection of creators' rights.

Writers also protested in March by publishing a book in the UK whose pages were blank, although the authors' names were listed on it. The goal of this act was to demonstrate that if AI companies are allowed to use others' work without permission, the future of real authors and their creativity will be threatened.

A similar protest was observed in the music industry: over a thousand artists released an album containing only the sounds of empty studios, as a reaction to the use of artists' works in training AI models.

A new front of protest has opened up in the US related to AI data centers. AI models require colossal computing power, leading to the construction of giant data centers. These centers require a very large amount of electricity and water, causing concern among local residents. Some residents suffer from noise emanating from the data centers, while others suffer from their side effects.