A US federal judge has ruled that a tech company, Anthropic, made “fair use” of copyrighted works by authors Andrea Bartz, Charles Graeber and Kirk Wallace Johnson to train its large language model (LLM) without permission. However, the judge also found that Anthropic’s storage of over 7m pirated books infringed on the authors’ copyrights, leading to a trial in December to determine how much the company owes for the infringement. The ruling has sparked debate among copyright owners and AI firms, with some hailing it as a significant decision that recognizes the transformative nature of AI training, while others argue that it sets a worrying precedent for the use of copyrighted material in AI development.
Source: https://www.theguardian.com/technology/2025/jun/25/anthropic-did-not-breach-copyright-when-training-ai-on-books-without-permission-court-rules