Sarah Silverman's lawsuit against OpenAI will move forward with the dismissal of certain claims from her legal team. The comedian continued OpenAI and Meta in July 2023, claiming they trained their AI models on his books and other works without consent. Bloomberg reported Tuesday that the unfair competition aspect of the trial will continue. Judge Martínez-Olguín gave the plaintiffs until March 13 to amend the complaint.
U.S. District Judge Araceli Martínez-Olguín on Monday dismissed parts of the complaint filed by Silverman's legal team, including negligence, unjust enrichment, DMCA violations and charges of constructive infringement. The main claim of the case remains intact. It alleges that OpenAI directly infringed copyrighted material by training LLMs on millions of books without permission.
OpenAI's motion to dismiss, deposit in August, did not address the case's main copyright claims. Although the suit will continue, the judge suggested that the Federal Copyright Act could preempt the suit's remaining claims. “As OpenAI does not raise preemption, the Court does not consider it,” Martínez-Olguín wrote.
The American justice system has not yet determined whether train large AI language models on copyrighted works falls under the fair use doctrine. Last month, OpenAI admitted in court filing that it would be “impossible to train today’s leading AI models without using copyrighted material.”
The outcome of Silverman's OpenAI hearing is similar to that in San Francisco in November, when Silverman's claims against Meta were also cut to major copyright infringement claims. At that session, U.S. District Judge Vince Chhabria called some of the plaintiffs' dismissed claims “absurd.”
Other groups sue OpenAI for alleged copyright violations include The New York TimesA collection of non-fiction authors (a group that grew up after the initial trial) and The Authors Guild. The latter filed her complaint alongside the authors George RR Martin (Game Of Thrones) and John Grisham.