Anthropic, the company behind the Claude AI assistant, has agreed to a $1.5 billion settlement in the Bartz v. Anthropic case after using pirated books from Library Genesis and Pirate Library Mirror to train its language models. The settlement covers approximately 500,000 copyrighted works and will pay qualifying US-registered rightsholders roughly $3,000 per title. A US District Court ruled that downloading copyrighted works from piracy platforms constitutes infringement. Anthropic is required to destroy the pirated datasets used in training. The Irish Writers’ Union has reached its own arrangement with Anthropic and sees the case as a potential precedent for European copyright enforcement against AI companies.
Source: Read the original article
