Judge dismisses authors copyright lawsuit against Meta over AI training

In a significant ruling, a federal judge has dismissed a copyright infringement lawsuit filed by a group of authors against Meta Platforms, the parent company of Facebook. The lawsuit alleged that Meta unlawfully utilized the authors' works to train its artificial intelligence systems. U.S. District Judge Vince Chhabria issued the ruling, marking the second time in a week that a San Francisco federal court has rejected major copyright claims from authors against the burgeoning AI sector. Chhabria stated that the 13 authors involved in the case had "made the wrong arguments," leading to the dismissal. However, he clarified that this verdict does not imply that Meta's use of copyrighted materials is legally acceptable. Prominent writers including comedian Sarah Silverman, Jacqueline Woodson, and Ta-Nehisi Coates were part of the plaintiff group but did not immediately respond to requests for comment. Meta also refrained from commenting on the ruling at this time. Judge Chhabria emphasized that while the ruling reflects the inadequacy of the plaintiffs' arguments, it does not address the legality of Meta's practices in relation to copyright. This decision follows another ruling from U.S. District Judge William Alsup regarding AI company Anthropic, which was found not to have violated laws by training its AI chatbot on copyrighted texts, although it must still face trial for allegedly acquiring those texts from pirated sources. Alsup's judgment indicated that the process of AI learning from various works could qualify as "fair use" under U.S. copyright law due to its transformative nature, a perspective that Chhabria criticized for overlooking potential market harm. In their court filings, the authors contended that Meta engaged in extensive copyright infringement by sourcing their books from online piracy repositories to enhance its generative AI system, Llama. The plaintiffs argued that Meta should have compensated them for their literary works. In response, Meta's legal team asserted that U.S. copyright law permits the unauthorized replication of works for transformative purposes, arguing that the output generated by its AI does not replicate the original texts. They maintained that Llama is not capable of producing the exact works from which it was trained. The lawsuit had significant implications, compelling Meta CEO Mark Zuckerberg to provide testimony and revealing internal discussions within the company regarding the ethical implications of using pirated materials. The authors' attorneys pointed out that Meta's awareness of the risks associated with utilizing copyrighted works from illicit databases was evident, as it escalated the issue to senior executives for approval. Judge Chhabria noted that, despite the dismissal, the ruling's ramifications are limited to the 13 authors involved and do not affect others whose works may have been used in Meta's AI training. This case underscores the ongoing legal complexities and ethical debates surrounding the use of copyrighted materials in the development of artificial intelligence technologies.