Amazon Adds Self-Publishing Restrictions to Counter Flood of AI Books

To curb the unauthorized use of AI in publishing chatbot-written books claimed to be by human authors on its Kindle platform, e-commerce behemoth Amazon has released new guidelines requiring publishers to disclose the use of AI in content submitted to the Kindle Direct Publishing (KDP) platform.

The new stipulations require publishers to inform Amazon about AI-generated content, including text, cover art, interior images, or translations while publishing or updating an existing book. On Wednesday, Amazon imposed limits on the number of titles that can be submitted and added AI-related questions to the KDP Publishing Process earlier this month.

While Amazon said publishers are required to disclose fully AI-generated content, it said there is no requirement to disclose AI-assisted content. According to the company, that refers to situations in which an individual creates the content themselves and employs AI-based tools to brainstorm ideas, edit, refine, error-check, or otherwise improve that content, whether it be text or images.

Amazon says publishers are responsible for verifying that all AI-generated or AI-assisted content adheres to its content guidelines and applicable intellectual property rights.

Generative AI has taken the world by storm since the launch of OpenAI’s ChatGPT late last year. With generative AI, users can create text, images, music, and videos from prompts entered into an AI chatbot. While that has made it easier for people to create all sorts of content, it has also led to increased copyright infringement, deepfakes, and privacy questions.

“Amazon is constantly evaluating emerging technologies and is committed to providing the best possible shopping, reading, and publishing experience for our authors and customers,” Amazon spokesperson Ashley Vanicek told Decrypt in an email. “All publishers in the store must adhere to our content guidelines, regardless of how the content was created.”

According to Vanicek, Amazon invests “significant time and resources” to ensure its policies are followed, saying that the company removes books that do not adhere to them.

“While we allow AI-generated content, we will reject or remove AI-generated content that we determine creates a disappointing customer experience,” she said.

The policy update comes a month after a controversy surrounding AI-generated books. Titles that were claimed to have been written by journalist and author Jane Friedman—but actually were not—were found on the Amazon Kindle website.

The books, including “Publishing Power: Navigating Amazon’s Kindle Direct Publishing,” which were said to have been written by ChatGPT, even found their way to Friedman’s Good Reads profile.

Friedman approached Amazon about the books and asked for them to be removed, but she claims that the retailer refused to take the books down due to Friedman not owning the trademark on her name. The books were removed from the site after the Authors Guild offered to step in on Friedman’s behalf.

The Authors Guild declined Decrypt’s request for comment on the new guidelines.

On Wednesday, several high-profile writers and authors—including John Grisham and George R.R. Martin—joined a class action lawsuit filed by the Authors Guild against ChatGPT creator OpenAI, alleging the AI developer violated copyright laws by feeding their works into the chatbot’s train model. Comedian Sarah Silverman filed a separate lawsuit against OpenAI earlier this year.

“This case is merely the beginning of our battle to defend authors from theft by OpenAI and other generative AI,” Authors Guild President Maya Shanbhag Lang said in a statement.