I’ve written recently about whether and how AI-generated content might become copyrightable, but this skirts the issue of whether the tools themselves are committing copyright infringement. Most advocacy groups for artists and creators say without a doubt, yes, it is infringement, and rights holders must be compensated for training the model. Lawsuits are already making their way through the courts.
However, the case isn’t a slam dunk. Authors Guild executive director Mary Rasenberger has said on multiple panels that she’s not optimistic these cases will be decided in favor of artists. The example often used is the Google Books scanning suit, where Google scanned millions of print books in order to make the books searchable. After a years-long battle in court, Google prevailed because its use was deemed fair. The output, search snippets of books, was considered non-infringing and the use transformative, even though it might have required infringement (wholescale copying) to get there. And so now anyone can search Google Books or use Talk to Books.
The parallels to ChatGPT are obvious: The input stage certainly looks like infringement, but the output may not be considered infringing by the courts. Certainly the narrower the question you pose to ChatGPT, the more likely it will paraphrase something that’s from a single source (e.g., ask it for your author bio; it will probably rely on something you wrote on your website or at a retailer). But paraphrasing from a single source is considered fair use, and all authors benefit from fair use when writing their own works.
For the sake of argument, let’s say that, whether pre-emptively or due to a court order, companies like OpenAI must make restitution to rights holders or pay for licensing works in its corpus. The Authors Guild and others have suggested a collective licensing solution that would compensate creators whose works are used to train generative AI. I wonder if the payments that result would amount to anything more than pocket change for the average recipient. Sure, you’ve made Big Tech pay, and congratulations on that. But would this really amount to meaningful support of arts and culture?
I propose a more radical idea. Ask those who profit from these AI models to put some of those profits into a fund that supports the arts and humanities. Then put some responsible people or organizations in charge of it, and create programs that make it easier for artists of all kinds to earn a living from their work.
Jane Friedman has spent her entire career working in the publishing industry, with a focus on business reporting and author education. Established in 2015, her newsletter The Bottom Line provides nuanced market intelligence to thousands of authors and industry professionals; in 2023, she was named Publishing Commentator of the Year by Digital Book World.
Jane’s expertise regularly features in major media outlets such as The New York Times, The Atlantic, NPR, The Today Show, Wired, The Guardian, Fox News, and BBC. Her book, The Business of Being a Writer, Second Edition (The University of Chicago Press), is used as a classroom text by many writing and publishing degree programs. She reaches thousands through speaking engagements and workshops at diverse venues worldwide, including NYU’s Advanced Publishing Institute, Frankfurt Book Fair, and numerous MFA programs.
