Two federal judges ruled against authors’ copyright lawsuits, allowing Meta and Anthropic to use books without permission to train their artificial intelligence systems. U.S. District Judges Vince Chhabria and William Alsup sided with the companies, citing “fair use” of copyrighted materials as a defense.
Meta Platforms won its case, with Judge Chhabria stating that authors did not present enough evidence to prove copyright infringement. The judge noted that Meta’s use of AI was not necessarily lawful, but the arguments presented by the plaintiffs were incorrect.
Anthropic’s ruling was similar, with Judge Alsup citing “fair use” as a defense and ordering a trial in December to determine how much Anthropic owes for allegedly infringing on the copyrights of Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson. The judge found that Anthropic’s copying and storage of over 7 million pirated books infringed on the authors’ copyrights.
The “fair use” argument is a key defense for tech companies, as it allows them to make use of copyrighted material without permission in order to create new, transformative content. AI companies argue that their systems can generate fair use content, and that paying copyright holders could stifle the growing AI industry.
Copyright owners claim that AI companies are unlawfully copying their work to generate competing content that threatens their livelihoods. However, judges have expressed sympathy for this argument, noting that generative AI has the potential to flood the market with endless content using a fraction of the time and creativity required to create it.
The ruling allows Meta and Anthropic to continue using books without permission to train their AI systems, but sets a precedent for future cases involving copyright infringement.
Source: https://www.foxbusiness.com/technology/judges-side-major-ai-companies-models-may-train-books-without-author-permission