Artificial Intelligence (AI) Policy

As technology advances, new techniques in publishing arise, and with the rapid development of AI and its influence on the future it is necessary that Andrews University Press officially maintain a policy addressing the appropriate use of AI in the writing process. Currently generative tools are text-completion tools, not information tools, and while this might change, it is important that authors recognize where AI is being used and what is appropriate use of AI as they prepare manuscripts to submit for publication.

Due to rapid changes, this policy will be updated regularly.

Andrews University Press recognizes that the appropriate use of AI tools has the potential to enhance research and its presentation. In order to maintain a standard of integrity, the Press asks that authors who wish to have a manuscript considered for publication should follow certain guidelines.

Andrews University Press recognizes that generative AI tools are based on Large Language Models (LLMs) that currently (2024) are associated with blatant and open infringement practices, and thus it is in the best professional interest of researchers to continue to commit to maintaining standards that recognize sources of their information.

Andrews University Press will not knowingly publish any manuscripts in which authors used generative AI in violation of the following best-practice guidelines:

• AI should only be used to support rather than replace research and is appropriate for brainstorming, refining an author’s ideas, and editing but must not serve as the primary source for data or the primary writing agent.
• Appropriate use of AI includes spell-checking, grammar, or syntax-related elements.
• AI cannot legally serve as a co-author (since AI is not a moral agent and cannot be held liable for the content of what it generates); therefore, authors must write the entire manuscript themselves, maintaining a unified voice.

Any AI-generated text (even if rewritten), any AI-generated outline followed by the author, or any image, graph, table, or the like generated by AI must be disclosed to the publisher and the reader. Further, any AI-generated data must be thoroughly reviewed, fact-checked, and sourced by the author. Failure to disclose AI influence may result in the rejection of the manuscript and will void any contractual obligations Andrews University Press may have with the author. In this way, failure to disclose AI influence will be treated in the same way as plagiarism.

In return, Andrews University Press pledges to not knowingly submit any of its published material to be used for LLMs without express written permission by its affected authors. At the same time, it is important for authors to recognize that submission of a manuscript for an AI grammar check is also a submission to an LLM.