University of Michigan Press supports the ethical adoption of technology, including generative AI, that can advance our mission; to maximize access to the expertise that our authors bring. In April 2025, we developed a set of "artificial intelligence practices" to shape our engagement with the increasingly widespread use of Generative AI. These practices include the expectation that we will license for Large Language Model training only the books of contactable authors who have opted-in. The interface between publishers, authors, and technology companies in regards to AI is an evolving area and we expect to update this section of the website as the environment involves.

1. When we use GenAI, we will use it to augment human labor rather than replace it.
2. We commit to reviewing GenAI work for accuracy before sharing it externally.
3. When we license materials to third parties for use in their training materials, we will be guided by the rights of our authors to credit and (where relevant) compensation.
4. We will offer our authors the option to opt-in/opt-out of having their work licensed for AI.
5. We will include information in our publications about when and how AI has been used.
6. We will require that authors/creators/reviewers are transparent when they use AI.
7. We will be careful not to undermine the integrity of our publications when employing GenAI.
8. As the manifold ethical issues around GenAI continue to evolve, we commit to continuing to research and discuss strategies for engaging with them in publishing work.
9. We will continue to review and develop these practices as technological, legal, and social considerations evolve.

For further questions not answered above, please feel free to contact us at um.press.rights@umich.edu