My general practice is to file for copyrights on behalf of my clients at the end of the publishing process. With the advent of Artificial Intelligence (AI), it poses a lot of questions as to what exactly is copyrightable if AI has been used.
When you upload files now for publishing, the distributors such as IngramSpark and Amazon ask questions about whether AI was used in the creation of the book or cover. Then, they ask more detailed questions such as, how much of it was used: was it a little bit or a lot, was it edited somewhat or heavily, etc. That leaves a lot of gray area around these questions. How do you determine “a lot”? Some of them also ask what AI programs were used. I would imagine this is for their own safety if lawsuits ever emerge against any particular AI platform.
So, back to copyrighting. If you used AI images in a children’s book, for example, then the software used by the copyright office can determine that it indeed has been used. That means it is NOT copyrightable. When you apply for the copyright, you must check the box on the application to LIMIT the copyright to text only and exclude the images.
Earlier this year, the U.S. Copyright office ruled that work generated with Artificial Intelligence CAN be copyrighted but the determination to whether it can or not falls as to just how much of the work includes “human input.” It is supposed to be “substantial.” Again, that raises several questions as to the interpretation of “substantial.” Stay tuned for more legal wrangling.