In the potentially the first huge payment of creations, whose work was used to train AI systems, Anthropic reached a contract for paying “at least” stunning $ 1.5 billion plus interest to settle a collective claim. The amount breaks down to smaller payments, which have about USD 3000 for a book or work. The lawyers of the plaintiffs said that “it is believed that this is the largest publicly reported revival in the history of American authorities.”
“This result is unusual,” legal application countries.
The settlement is subject to approval by the court, and the trial will take place on September 8. According to the press release, the final amount may be higher, because about 500,000 works will probably be paid, but if the sum is higher, Anthropic will pay an additional 3000 USD for work, and all depends on the number of claims submitted. As part of the settlement, Antropic must also destroy the original downloaded files and any copies.
“The settlement only releases claims based on previous files – does not give anthropia license or permits for future AI training and does not disclose any claims that arise after August 25, 2025.” – stated in a communication.
A breakthrough decision is part of the annual saga. In August 2024, three authors – Andrea Bartz, Charles Graeber and Kirk Wallace Johnson – alleged This Anthropic “built activities worth many billion dollars, stealing hundreds of thousands of copyright protected books.” In the summer, the federal judge gave Anthrop a slight victory, ruling that the company has the right to train its AI models in the scope of purchased books. But the judge also said that Anthropic would have to face a separate trial for the alleged operate of pirate books. In July, a federal judge in California ruled that the authors could bring a collective lawsuit representing all US writers, whose work allegedly came from the libraries of pirate books taken by Anthropic. And at the end of last month, Anthropic said that he had decided a collective claim – but so far we did not know the conditions.
“In June, the District Court issued a breakthrough decision on the development of AI and copyright, stating that Anthropik’s approach to training AI models is allowed,” said Kalen Sridhar, deputy advisor to Anthropic, deputy advisor to general Anthropik, The Verge in a statement. “Today’s settlement, if approved, solves the remaining older claims of the plaintiffs. We remain involved in the development of safe AI systems that help people and organizations expand their possibilities, develop a scientific discovery and solve complex problems.”
The settlement is particularly noteworthy in times of approaching court trials against AI in the case of infringement of copyright from the media, platforms, creations and companies – and at the same time an increasing number of partnerships with the same AI companies from companies and media willing to provide data to training AI systems to obtain a piece of Pie.
In recent years this is not the only legal action. This summer, Reddit filed a lawsuit against Anthropic, claiming that AI bots have gained access to Reddit over 100,000 times since July 2024, after Antropic said that he blocked them. And in 2023, Universal Music sued Anthropic about “systematic and universal violation of their protected right songs.”
In this matter, the authors and right -wing can visit Anthropiccopyrightsetetlement.comThe issue stated that the website, which “gives potential class members the option to provide contact information for class advisers.”
“In the coming weeks, and if the court initially approves the settlement, the website will provide a full and easy list of all works covered by the settlement and information for class members on their options and settlement rights,” says the message.
0 Comments