Anthropic to pay authors $1.5B to settle AI copyright lawsuit
What's the story
A US District Judge has approved a $1.5 billion settlement between artificial intelligence (AI) company Anthropic and authors whose works were allegedly used without permission to train chatbots. The settlement was given preliminary approval by Judge William Alsup in a San Francisco federal court on Thursday. Anthropic will pay authors and publishers about $3,000 for each of the books covered by the agreement.
Legal proceedings
Settlement fair but distribution complicated: Judge Alsup
The settlement comes after Anthropic was accused of illegally pirating nearly half a million books to train its AI chatbots. Judge Alsup said that while this is a fair settlement, distributing it to all parties will be complicated. Justin Nelson, an attorney for the authors, confirmed that about 465,000 books are on the list of works pirated by Anthropic.
Industry response
Publishers group calls it major step forward
The American Association of Publishers has welcomed the settlement as a major step toward holding AI developers accountable for copyright infringement. Maria A. Pallante, president and CEO of the publisher group, said Anthropic's case is not unique and every other major AI developer has trained their models on authors' and publishers' works sourced from notorious infringing sites.
Company statement
Anthropic satisfied with preliminary approval of settlement
Anthropic has expressed its satisfaction with the preliminary approval of the settlement. Aparna Sridhar, deputy general counsel at Anthropic, said the decision will let them focus on developing safe AI systems. She added that the court's landmark June ruling that AI training constitutes transformative fair use remains intact and this settlement simply resolves narrow claims about how certain materials were obtained.