The article discusses the ethical and legal implications of generative artificial intelligence (AI) systems relying on copyrighted works for training. It highlights that advanced AI models like those developed by OpenAI, Anthropic, and xAI have been trained using massive repositories of literary, artistic, and scientific works protected by copyright, often obtained through web scraping without specific licenses. While these companies promote a future where human labor becomes obsolete, they rely heavily on the intellectual contributions of humans whose work forms the foundation of their AI technologies. The article questions whether these AI firms should compensate authors and rights holders for using their content, noting that the European Union allows exceptions for text and data mining under certain conditions, but respects the opt-out rights of creators.
Ocena pristranskosti (Levo): The article emphasizes the moral and legal debt owed to human creators whose work fuels AI development, critiques the lack of acknowledgment by AI companies, and frames the issue as an ethical obligation rather than a purely economic transaction. This framing aligns with progressive values that seek
