On October 10, 2025, Law 132/2025 entered into force, marking Italy’s first legislative response to Regulation (EU) 2024/1689 on artificial intelligence, the new European legal framework for AI. The EU regulation – adopted last June – represents the first attempt to systematically regulate the development and use of artificial intelligence, using a risk-based approach.

The European legislator has classified AI systems into four categories (from unacceptable risk to minimal risk), imposing strict obligations on high-risk systems, such as those used in healthcare, justice, critical infrastructure, labor, and education. This category also includes systems used in the notarial sector and, more generally, in the administration of justice, where they affect fundamental rights or public trust.

The Italian legislator has introduced a specific principle which establishes that AI may only be used as a supportive tool for professional activity, with human activity prevailing, and requires informing the client. The Italian Consiglio Nazionale Forense has therefore prepared a guideline that lawyers can use in their interactions with clients and assisted parties (the text is available at: https://shorturl.at/NZpZC).

For notaries, this means that AI cannot replace personal intervention in the delicate phase of investigating the parties’ intentions. The new law therefore outlines paths for compliant use, with particular attention to professionals’ obligations and responsibilities.

While the law seeks to protect the central role of human activity, it also confirms the need to ensure transparency, security, and reliability in AI use, without introducing additional constraints beyond the European regulation.