Authors vs. AI Scraping: Round Three
I have written two articles about how AI companies are legally/illegally using books to train their AI models and the fight that authors are waging against this practice. It is getting so bad that authors are protesting:
For a small-time author like me, this is a losing battle, but I may be getting a new ally. Sony recently developed technology that identifies when an AI company illegally trains its model with copyrighted Sony music. This software is far more complicated than identifying when a book’s pages have been used, so similar software will likely be available to authors soon.
What do I think about this possibility? My first reaction is that this tool would be amazing, but I would be quickly reminded that I do not have the means to fight a huge corporation. Therefore, unless laws are enacted that help me fight the system, being given proof that AI has scraped my precious words would only bring disappointment. “Dang, some AI company mooched off my hard work.”
Yet something else is going on. In my teens, I liked music but did not have the money to buy records. The next best thing was recording music from the radio, and I spent hours making mixtapes. Musicians/record companies were aware of this practice, but they gave kids a pass because they knew the kids would grow up, get jobs, and buy records. It certainly was true for me, and I still have a giant CD collection wasting space in my den. Yet, I would argue that AI scraping is different.
Teenage Bill making a copy of a song off the radio was a single person doing a single thing. AI machine learning is an uncaring beast doing the bidding of a large corporation, which uses the results to make money. Meaning this machine is impersonal, and authors will never be able to see the enemy. This all reminds me of the Metallica vs. Napster lawsuit:
Yet something else is going on. In my teens, I liked music but did not have the money to buy records. The next best thing was recording music from the radio, and I spent hours making mixtapes. Musicians/record companies were aware of this practice, but they gave kids a pass because they knew the kids would grow up, get jobs, and buy records. It certainly was true for me, and I still have a giant CD collection wasting space in my den. Yet, I would argue that AI scraping is different.
Teenage Bill making a copy of a song off the radio was a single person doing a single thing. AI machine learning is an uncaring beast doing the bidding of a large corporation, which uses the results to make money. Meaning this machine is impersonal, and authors will never be able to see the enemy. This all reminds me of the Metallica vs. Napster lawsuit:
Also, I feel for large corporations, because I worked for a company that developed products. There, I spent long hours designing, testing, and manufacturing something that was sold. Did I cross any lines in the process? I consider myself a highly ethical person, but I cut corners in the name of speed, saving money, and improving quality. So, yes, I did things that could be considered wrong.
My point is that I took great pride in my work, and I am sure the people who worked for these AI corporations do as well. Thus, when they needed training material, they turned to books. While I do not like it, I understand that the only other option was to spend billions paying authors to write books exclusively for AI training, and this was out of the question.
Still, when this software/service arrives, I am sure these large AI corporations will get a torrent of lawsuits. Would I join one of these class actions? Of course, but I would not expect to get any money because I was part of three cell phone class-action lawsuits that paid me less than $10. I cannot imagine that a small-time author like me could expect more than a few cents. Dang, I dislike ending my articles on a down note.
You’re the best -Bill
March 25, 2026
My point is that I took great pride in my work, and I am sure the people who worked for these AI corporations do as well. Thus, when they needed training material, they turned to books. While I do not like it, I understand that the only other option was to spend billions paying authors to write books exclusively for AI training, and this was out of the question.
Still, when this software/service arrives, I am sure these large AI corporations will get a torrent of lawsuits. Would I join one of these class actions? Of course, but I would not expect to get any money because I was part of three cell phone class-action lawsuits that paid me less than $10. I cannot imagine that a small-time author like me could expect more than a few cents. Dang, I dislike ending my articles on a down note.
You’re the best -Bill
March 25, 2026
