A federal judge recently made a significant ruling in a lawsuit filed by Sarah Silverman against Meta over the unauthorized use of authors’ copyrighted books to train its generative artificial intelligence model. The judge dismissed most of Silverman’s claims, marking a significant win for AI firms in the ongoing legal battle over intellectual property rights and the use of copyrighted material for AI training.
In his ruling, U.S. District Judge Vince Chhabria offered a full-throated denial of one of the authors’ core theories that Meta’s AI system is itself an infringing derivative work made possible only by information extracted from copyrighted material. Chhabria stated that there was no way to understand the AI models as a recasting or adaptation of any of the plaintiffs’ books. Another argument that every result produced by Meta’s AI tools constitutes copyright infringement was also dismissed unless there was evidence that any of the outputs could be understood as recasting, transforming, or adapting the plaintiffs’ books.
Despite these dismissals, Meta didn’t move to dismiss the allegation that the copying of books for purposes of training its AI model rises to the level of copyright infringement. This suggests that this aspect of the claim may still be viable and could have implications for similar cases in the future.
This ruling builds upon findings from another federal judge who oversaw a lawsuit from artists suing AI art generators over the use of billions of images downloaded from the internet as training data. In that case, the judge delivered a blow to fundamental contentions in the lawsuit by questioning whether artists can substantiate copyright infringement in the absence of identical material created by the AI tools. These rulings indicate that the legal disputes surrounding AI-generated outputs and their use of copyrighted material for training are complex and have significant implications for creators, AI companies, and the future development of AI technologies.
The issues presented in the litigation could decide whether creators are compensated for the use of their material to train human-mimicking chatbots that have the potential to undercut their labor. AI companies maintain that they don’t have to secure licenses because they’re protected by the fair use defense to copyright infringement. However, the plaintiffs have argued that the entire purpose of Meta’s AI model is to imitate copyrighted expression and that the entire model should be considered an infringing derivative work.
The judge’s ruling suggests that plaintiffs across most cases will have to present evidence of infringing works produced by AI tools that are identical to their copyrighted material. This potentially presents a major issue because they have conceded in some instances that none of the outputs are likely to be a close match to material used in the training data. This means that under copyright law, a test of substantial similarity is used to assess the degree of similarity to determine whether infringement has occurred.
Some of the other dismissed claims include those over unjust enrichment and violation of competition laws, which the judge found to be preempted by the surviving claim for copyright infringement. Meta did not immediately respond to a request for comment following the ruling.
In addition to the lawsuit against Meta, Silverman also joined a class action against OpenAI accusing the company of copyright infringement. The case has been consolidated with other suits from authors in federal court.
The judge’s ruling and the ongoing legal battles between creators and AI companies highlight the complex, evolving nature of intellectual property law in the context of technological advancements like AI. These disputes will continue to shape the legal landscape, and their outcomes will have far-reaching implications for the future development and use of AI technologies.