Parisian Phoenix Publishing

In the car this morning, I listened to a Wall Street Journal podcast about the recent lawsuit between authors and Anthropic AI. A judge recently ruled that Anthropic could use authors’ work to train its AI model under fair use if the works were acquired in such a way that compensated the author. Originally, the lawsuit included seven million published works, but the courts whittled that down to about 500,000 by eliminating duplicates and checking with the copyright office.

The lawsuit used the copyright registrations on file with the United States Copyright Office to determine if the author were entitled to compensation from the lawsuit. If you are a writer who does not or does not intend to register your works with the copyright office, read that again.

In general, in my understanding and I am no lawyer, copyright lawsuits are expensive and tricky. As soon as you place words on the page, as simply as I type these words here, those words belong to you and fall under the protection of copyright laws, plain and simple. No official registration required.

If you want to enforce your copyright— such as in this potentially class action suit against Anthropic— your damages will be higher and your case easier to prove if you have a copyright registered.

A lot of authors I know chose not to spend the $50 or $65 on copyright registration but in the current world where AI model training is a “Wild West,” copyright registration might be a wise investment. I believe it will only get more difficult to prove if something is indeed “yours.”

Jane Friedman offered her analysis in her paid newsletter, but provided a free link here.