Publishing is a collaborative human endeavor that is bound by trust between the author and publisher. This trust extends to every partner we engage throughout a book鈥檚 lifecycle鈥攑eer reviewers, freelance editors and designers, printers, distributors, librarians, booksellers, media, and readers.
One of the many ways in which we invest in an infrastructure of trust is the protection and stewarding of copyrights, the form of intellectual property (IP) that protects original works. Copyrights are an essential tool for communicating the authenticity of a work and most notably in an era of AI, they are a trusted indicator that human creativity has generated that work.
While a powerful new tool in many spheres, AI also profoundly challenges copyright, integrity of content, and trust. Many LLMs were initially trained on pirated content, 快色直播 books included, as rather than license content from publishers, big tech firms chose (with intent as legal cases are documenting) to use IP from pirate sites like Anna鈥檚 Archive, PiLiMi, and LibGen to train LLMs.
As generative AI tools become more widely available and powerful, we need to continuously adapt our assumptions and understanding about its responsible use. For example, PUP has made the decision not to license Press books for use in LLM training, as this training is not grounded in trust and it does not adhere to what we believe are licensing essentials: copyright, consent, credit, and compensation to our authors and rights holders.
We are learning every day how to responsibly engage AI in our publishing, but even with our collective best intentions, the tools widely available necessitate critical thinking and added vigilance, as ethical guardrails have not kept pace with technological development. To ensure human authored and human centered publishing remains resilient, we need your support to help us to protect copyright.
What does this mean?
If you are reviewing a manuscript, evaluating a proposal, considering an endorsement or blurb, admiring a cover we have designed, or engaging with a publication as a reviewer, or as a reader, first and foremost, we thank you! Second, we ask that you respect the rights of the creators. In the conventions of consent, copyright, and credit, please refrain from uploading any Press materials in whole or in part, into public or private AI systems or chatbots. This includes manuscripts, proposals, sample chapters, PDFs, illustrations, cover designs, marketing materials, and any other unpublished or published content shared for your review and engagement.
Why does this matter?
When unpublished or published material is uploaded to an AI platform, it may become subject to terms and processes that are outside the control of the author, reviewer, or publisher. Even when AI platforms offer assurances about how content is handled, the act of sharing a copyrighted work with a third party violates copyright conventions and creates a cascade of risks, including undoing the impact of the work we undertake every day to protect copyright so these creative works can be engaged with, and credited, in perpetuity.
For peer reviewers, there is an additional concern. The review process depends on confidentiality, independent judgment, human expertise, and deep engagement with the work itself. Delegating that engagement to an AI tool undermines the purpose of peer review and weakens the trust and expertise that make scholarly publishing uniquely valuable in the information ecosystem.
For endorsers and blurb writers, the value of an endorsement comes from genuine interaction with a book. Readers, authors, and publishers all rely on the understanding that a recommendation reflects a person's own assessment and response. AI may assist with editing prose, but it should not replace the reading, thinking, and judgment that give an endorsement its meaning.
For accessibility needs, we offer multiple formats directly and with partners. We can also offer guidance on how these technologies should鈥攁nd should not鈥� be used when working with us throughout the book publication lifecycle.
We recognize the inclination toward using AI tools to save time鈥攐ne of our most precious resources鈥攁nd we understand the appeal of quickly summarizing a manuscript, identifying themes, gathering endorsement language, or generating review notes or an index. We trust positive intentions. However, these good intentions have material consequences.
Our commitment is straightforward: we are working to protect authors' rights, preserve the value of original creative work, and maintain the human relationships that sit at the heart of publishing. Each time unpublished content is uploaded to an AI system without permission, it becomes harder to protect creators and their work.
We are grateful to our authors, reviewers, endorsers, and publishing partners for helping us uphold those standards. Protecting creative work is a shared responsibility, and one with implications for the future of book publishing. We thank you for being part of that effort.