To enhance its AI assistant Claude, the artificial intelligence firm Anthropic embarked on an unprecedented initiative. The company acquired a vast collection of physical books, converted them into digital format using destructive scanning methods, and subsequently discarded the original paper copies. By legally purchasing the books and mimicking Google's strategy for digitizing literature, Anthropic sought to bypass traditional authorization processes. Despite courts ruling its scanning practices as fair use, initial instances of utilizing pirated e-books ignited controversy. In contrast, other firms opt for non-destructive scanning techniques. Anthropic's approach has ignited widespread discourse concerning the intricate balance between copyright law and the training of AI models.