Sarah Silverman Lawsuit Against OpenAI Suffers Setback As Judge Trims Case
Top authors suing OpenAI over the use of their novels to train its artificial intelligence chatbot have hit a stumbling block, with a federal judge narrowing the scope of their case.
U.S. District Judge Araceli Martínez-Olguín on Tuesday evening dismissed a claim accusing the Sam Altman-led firm of unfair business practices by utilizing the works of authors — including Sarah Silverman, Paul Tremblay and Ta-Nehisi Coates — without consent or compensation to power its AI system.
More from The Hollywood Reporter
As Apple and OpenAI Grow Partnership, Studios Stand on Sidelines of AI Battle
OpenAI Sued by Center for Investigative Reporting as Media Takes Diverging Paths on Tech Giant
Time Inks Deal With OpenAI as Tech Giant Makes Major Publishing Moves
The writers’ primary claim for direct copyright infringement was left untouched.
In February, the court dismissed other claims for negligence, unjust enrichment and vicarious copyright infringement. It denied dismissal of the unfair competition law claim, but lawyers for the authors tweaked it after lawsuits from Silverman, Tremblay and Michael Chabon — all of whom originally brought their own class actions — were grouped together. OpenAI seized upon the changes for a second try at dismissal, which was challenged by the plaintiffs.
In the order, Martínez-Olguín not only found that the company is allowed to move to dismiss the claim but that the Copyright Act bars it. She said that the law “expressly preempts state law claims” relating to works “within the subject matter of copyright.”
The authors argued that the unfair business practice at issue was using their works to train ChatGPT without permission. But since the allegedly infringed materials are copyrighted books and plays, they cannot bring a state law claim, which the court concluded should be under the purview of copyright law.
“Thus, the UCL claim based on the copying of Plaintiffs’ Infringed Works falls squarely within the ambit of the Copyright Act,” wrote Martínez-Olguín, who noted that an opportunity to amend the claim would be “futile.”
OpenAI faces multiple fronts in the increasingly tense legal battle over the use of published works to train its human-mimicking chatbot. Trade groups, artists and publishers have also sued the company.
Best of The Hollywood Reporter