Authors sue OpenAI for copyright infringement

The inevitable has occurred: There’s a lawsuit against OpenAI’s ChatGPT that seeks class action status on behalf of authors. It’s already well established that OpenAI’s model, as well as others like it, has been trained on works protected under copyright. However, as these models continue to operate with little transparency, a lot of unknowns remain. We do know that entire books—some from Smashwords and others likely from pirate sites—have been scraped off the open web without permission.

Critical questions, likely to be debated for years to come, lie at the heart of this case: Is copyright being violated in the training of ChatGPT? Is ChatGPT itself an infringing derivative work? Are the outputs of ChatGPT infringing on copyright? Did OpenAI violate the DMCA? While authors may feel the answers are clear cut, these are complex issues.

The Authors Guild has issued a statement in support of the lawsuit and is “standing by to provide support if requested.” Their statement continues, “Using books and other copyrighted works to build highly profitable generative AI technologies without the consent or compensation of the authors of those works is blatantly unfair—whether or not a court ultimately finds it to be fair use.” Reading between the lines, one might conclude the Authors Guild believes this will be a challenging case to win. Thus far, the Guild has pursued legislation to protect authors in the AI era rather than sue the tech companies. That isn’t so surprising, given the failure of their suit against Google’s book scanning initiative, which was found to be fair use.

Meanwhile, not just authors are suing OpenAI. There’s another class action lawsuit involving all internet users, brought by another law firm, that wants to represent “real people whose information was stolen and commercially misappropriated to create this very powerful technology.”