In 2023 with the rise of Large Language Models (LLMs), Amazon put a publishing limit on three books per day per author (The Guardian). Without seeing a spike in publications at the time, they could clearly foresee the problems that were going to arise for the publishing industry.
We are now seeing a rise in AI imitations of published books appearing on Amazon, using the same titles and ideas from the original authors, often summarised to a short length. (Wired). Publishing summaries of books and passing them off as the real thing has been a problem since 2019 (Wall Street Journal), but the power of LLMs has seen their appearance increase dramatically.
These books breach Amazon’s terms of AI use and are not allowed to be published and sold on the platform. However, Amazon only removes content reactively after it has already gone live on the platform. This means they can be distributed and sold, before someone eventually flags the issue and has it removed.
Legally, it seems to be in a confusing space. Some copyright lawyers believe these summaries may be legal if they avoid word-for-word plagiarism, while others argue that summaries without analysis or new information added would be hard to defend from an IP standpoint.
This leaves the digital publishing world in a very confusing and problematic place, with writers and publishing experts concerned this trend will continue to grow, undermining original authors and become a lasting problem for the industry.