An Insightful AI Copyright Ruling

In my Substack blog, From the Pure Land, I write posts intended for people interested in Buddhism, other Eastern religions, and those simply exploring spiritual paths. I aim for posts that can be read in five minutes. You can check it out here.

I started this blog, Awakening for Regular Folks--on my author website--where I post shorter pieces primarily about spiritual topics, but sometimes on issues relevant to independent writers and publishers like me. This post is about what may prove to be a significant ruling in the legal battles between the print publishing industry and AI companies.

On June 23, Judge William Alsup of the United States District Court , Northern District of California, ruled that buying books and feeding their pages into computer storage for use in training large language models (LLMs) falls into the category of "fair use" under copyright law. It's OK when the training is to teach AI engines what good writing is like. The ruling would not apply if the purpose were to reproduce segments of the books.

Using pirated, not purchased, books does not qualify as fair use, Judge Alsup ruled, and the case--Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson vs. Anthropic PBC--will continue to trial on the pirating issue.

Many authors and publishers resist the idea that any form of copying the contents of books for use by LLMs is fair use. They want compensation, especially when one book bought at the regular price is copied for commercial use, but I agree with the judge, and I hope this AI copyright ruling stands. Perhaps one day, there will be a separate price structure for AI engines purchasing books, as there is now for libraries, but that system does not yet exist.