In September, the Authors Guild and 17 authors filed a class-action lawsuit against OpenAI for copyright infringement on their works of fiction, claiming their work had been used to train AI models without their consent. They’ve since filed an amendment to add Microsoft to the list of defendants. While a decision on the suit has yet to be made, many authors feel that their livelihoods hang in the balance.

Rebecca Ackermann wrote about what artificial intelligence means for authors and the book publishing industry in Esquire magazine. She joined “Marketplace” host Kimberly Adams to discuss her piece. The following is an edited transcript of their conversation.

Kimberly Adams: What are some of the concerns that authors have around AI?

Rebecca Ackermann: I think there are a few different ones. And I think the first one is really about confusion. AI as a whole doesn’t feel like it makes a lot of sense to people who are outside of the tech industry because it can mean a lot of different things. So, I think first off, authors aren’t really sure what the threat is from AI because they aren’t really sure what AI is, or what the companies who build it are trying to do. And then I think the other major concern is this replacement concern that we hear in other industries as well, that AI is going to replace authors. AI will be writing all the books. AI will be getting all the money. And authors who are already really squeezed financially are going to be out of luck.

Adams: Speaking of finances, what’s it like trying to make a living as an author these days?

Ackermann: It’s very challenging, and for writers as well, like myself. So, the Authors Guild, which is the oldest organization for writers in the United States, did a recent survey, and they found the median author income in 2022 was only $20,000, and only half of that was from book sales. So, it’s pretty rough right now for authors and writers financially to make a living from the work that they do.

Adams: You also get into how AI has affected the self-publishing industry. What’s going on there?

Ackermann: Yeah, so self-publishing has really been booming, for better or worse. In a lot of ways, a lot of authors feel that now they have access to putting books and work out there in a way that they haven’t before, which is very exciting. And I think on the other hand, there’s now a huge flood of content that’s out there for readers to pick and choose from and for them to spend their money on in different ways. In 2023, 2½ million books were self-published in the United States alone, and that’s a lot of books. And one of the biggest accelerators for self-publishing has been Amazon. The Kindle Direct Publishing platform that Amazon has supports authors in publishing their own works directly online to a distribution network that Amazon has set up.

Adams: Obviously, there are lawsuits ongoing by authors who feel like their work was inappropriately scraped to train AI models. And a lot of authors are worried about this. Is there anything that authors can do to protect their work from being used to train AI models at this point?

Ackermann: So, the Authors Guild recommends including AI clauses in book contracts for their work itself. So that is to say, this book cannot be used to train models. But it’s also for the work that’s surrounding the publication of their books. So, if there’s an audiobook, you can include a clause that says, “I want a human actor to record my audiobook.” So, there you can have protections both on how your content is used to go into AI, and what content comes out of AI. What’s sort of complicated about this conversation is that AI is both about what content is used to train AI, and then it’s also about the content that AI creates and how that content is used in place of human artists or in collaboration with human artists.

Adams: You are a writer yourself. And I wonder how all this makes you feel?

Ackermann: It’s a great question. So, I have a background in tech. I worked in user interface at a number of different tech companies before I moved into journalism. And the part for me that’s the most frustrating is the different languages that the tech world and the culture worlds are speaking, and how it’s very hard to get the two to agree on terminology and values. So, that disconnect between the two worlds I find to be the most frustrating because I want us to be able to find a way to incorporate technology in a way that also valorizes labor and humanity that goes into it.

Stories You Might Like

That best-selling romance book? It might be based on fan fiction

The economy and ethics of AI training data

What’s it like publishing your first book in a pandemic?

FTC sues academic publisher for deceptive practices

AI could automate over half of banking jobs

China could soon surpass U.S. in crucial emerging technologies, report warns