Abstract: The history of publishing can be seen as a history of publishing innovation. From the invention of the codex book form, the printing press, until digitization of book content, some innovations had huge impact on the development of human mankind. Has Artificial Intelligence（AI）the same potential to be a game changer in the future course of the publishing industry? When we analyze past developments and innovations, we can clearly see that many innovations in the past had intended consequences, but many also had consequences that were neither foreseen during the development process, not intended. Understanding that, we are looking at benefits of using AI techniques within academic publishing, but also at potential risks and downsides that can potentially harm the distribution with new, accurate, and relevant knowledge. We come to the conclusion that currently the benefits seem to outweigh the potential risk, but keeping in mind that innovation may also have unforeseen consequences, we need rules and guidelines in the process of integrating AI methods into the publishing process. These guidelines should include four principles: "Transparency", "Experiments", "End Goal" and "Participation".

Key words: Scholarly publishing, Artificial intelligence, Technical innovation