L’intelligenza artificiale e la proprietà intellettuale : le interferenze e la disciplina de iure e de iure condendo

The essay explores the complex intersection of artificial intelligence (AI) and intellectual property (IP) law, analyzing the legal implications concerning both the data inputs used to train AI systems and the outputs they generate. Regarding training data, the essay examines the increasing legal disputes over the alleged unauthorized use of copyrighted materials as well as the legality of web scraping for training purposes and the applicability of text and data mining exceptions. Concerning AI-generated outputs, the essay addresses the challenge of securing IP rights, particularly copyright, highlighting the ‘‘human author’’ requirement. Moreover the Author takes into consideration the need to identify IP protection for works deriving from the use of artificial intelligence as a tool in the hand of the human creator. The essay analyzes also the European Union’s regulatory response through the AI Act, the ongoing development of Codes of Conduct under the AI Act and the proposed Italian domestic law on AI, which considers the protection of works created with AI tools if they result from the author’s intellectual work. In conclusion, the essay underscores the necessity of a balanced approach that safeguards IP rights while fostering AI development. It introduces the concept of machine unlearning as a potential technique to address data privacy and IP issues post-training. The author suggests that guiding principles may be more suitable than rigid rules for navigating the evolving relationship between AI and IP.