A judge in the Southern District of New York denied OpenAI’s motion to dismiss a consolidated class action suit over ChatGPT. The suit combines lawsuits from authors including Michael Chabon, Ta-Nehisi Coates, Sarah Silverman, Kai Bird, and Victor LaValle, as well as the Authors Guild, alleging that ChatGPT’s outputs are similar to the authors’ work and constitute copyright infringement. Judge Sidney H. Stein determined that the plaintiffs’ argument is strong enough to go to trial. In trying to dismiss, OpenAI argued that ChatGPT summarizing books is the same as summarizing news articles, which “the court determined were not substantially similar to […]
AI
More Anthropic Settlement Documents and Answers Take Shape
In keeping with the court’s deadlines, plaintiffs’ counsel and Anthropic filed numerous additional materials on Tuesday. The attachments here show the proposed Claim Form and the proposed Long-Form Notice that will be sent to potential claimants explaining the settlement. Additionally, the attorneys filed their answers to Judge William Alsup’s 34 additional questions — many of which were already answered by the proposed plan of allocation filed on Monday. In those answers, we learn that the final corrected works list contains 482,460 works. They confirm that for any unclaimed work, those funds will be redistributed among all of the actual claimants. On […]
Author Plaintiffs Propose Claims Process for Anthropic Settlement, Including Default Options
The author plaintiffs submitted details to District Court Judge William Alsup proposing a notification and claims process in the hopes that the judge approves the $1.5 billion Anthropic copyright infringement settlement. Those plans, to be followed today by written answers to Judge Alsup’s 34 detailed questions, were filed ahead of the September 25 hearing, and supported by letters from a wide variety of author and publisher organizations and literary agencies. The filing notes, “The parties have worked around the clock since the Court’s initial preliminary approval hearing on September 8 to address the remaining issues the Court raised at that […]
Survey Shows Professionals Use AI, Despite Concerns
Results of a survey by the Book Industry Study Group show that almost half of North American publishing professionals are using AI in their work, but just about all of them have concerns about the technology. Of 559 participants, including publishers, librarians, manufacturers, individual consultants, and retailers, 45 percent say that they are experimenting with AI tools individually and 48 percent report that their organizations use it. Among individuals, the most common uses are for administrative tasks (23 percent) and data analysis (21 percent). Least common uses are in translation (4 percent) and rights and licensing management (2 percent). Organizations […]
Judge Alsup’s 17 Simple Settlement Questions
Further to Monday’s surprise hearing, when Northern California District Court Judge William Alsup declined to provide preliminary approval for the Anthropic copyright infringement settlement due in part to “important questions to be answered in the future,” he issued a new order on Wednesday. That order provides the parties with 17 questions to address with “complete and succinct written answers” by September 23, ahead of the next hearing on the settlement. Judge Alsup poses a variety of complex “scenarios,” regarding everything from disputed or conflicted claims on the same work; works with multiple authors; and works where a third-party like a […]
Authors Sue Apple Over AI Training
Authors Grady Hendrix and Jennifer Roberson have filed a class action lawsuit in the Northern District of California against Apple for copyright infringement using their books to train its LLM. The lawsuit asserts that Apple used the pirated dataset Books3 to train its language models, and that the company’s Applebot software scraped pirate sites to obtain copyrighted books. It also notes that Apple entered a licensing deal with Shutterstock to train its genAI tools, but not with authors. “Apple did not compensate creators for use of their copyrighted works and concealed the sources of their training datasets to evade legal […]