Against the use of AI: open letter from authors to publishers

Various authors have addressed major publishers in an open letter against the use of AI in content creation.

The open letter, titled “Against AI: An Open Letter From Writers to Publishers,” was published by Literary Hub on June 27, 2025, and is addressed to publishers such as Penguin Random House, HarperCollins, Simon & Schuster, Hachette Book Group, Macmillan, and all other American publishers. The authors see themselves at a turning point, emphasizing that their core mission as artists is to reflect the human experience. However, they warn that the world demands art that is fast, cheap, and on-demand, which could lead to a future where novels, biographies, poems, and memoirs are “written” by AI models that, by definition, cannot know what it means to be human, to suffer, to hunger, or to love.

The letter’s authors sharply criticize that the language used by AI bots in response to prompts is partly generated through the synthesis of artworks created by the signatory authors throughout their careers. This happens without their consent, without payment, or any other form of credit. The authors stress that their works are being stolen and used to train machines. They see this as another level of exploitation of their work, where someone else is paid for a technology built on their unpaid labor. Although AI is a powerful tool with potential for societal benefits, the authors do not see the replacement of art and artists by AI as such a benefit. They highlight that not only is their own work being stolen, but also that of their publishers, and that the jobs of editors, proofreaders, publicists, and publishers are at risk.

To counter this development, the authors call on publishers to take a clear stance and make a series of commitments. Specifically, they urge publishers to publicly commit to the following points:

• Publishers shall not, either openly or secretly, publish books written with the help of AI tools that have unlawfully used their authors’ works.

• Publishers shall not invent “authors” to market AI-generated books and shall not use pseudonyms to publish AI-generated books based on stolen works.

• No AI shall be used for the design of published books based on stolen works.

• No employee shall be replaced, in whole or in part, by AI tools.

• No new positions shall be created to oversee the production of AI-generated texts or artworks.

• Employee job descriptions shall not be changed to turn their positions into mere monitoring tasks for AI-generated content; for example, editors should continue to review the work of human authors.

• Only human audiobook narrators shall be hired, and no AI-tool-generated “narrators” based on stolen voices.

The petition associated with the open letter has since been signed by well over 1,000 people. However, the creators have already faced setbacks in court. In two cases, U.S. courts ruled in favor of Anthropic and Meta, granting them the right to train their language models using copyrighted works under the “Fair Use” doctrine, as long as they acquired these works through legally sound means.