Italy | Regulatory Framework
Status: in force
Effective: 2025-10-10
high Italy Law 132/2025 on artificial intelligence (principles, authorities, delegations, amendments)
Law 23 September 2025, No. 132 — Provisions and delegations to the Government on artificial intelligence
I. Regulatory Summary
High impact across public sector, regulated actors, and the legal system: it establishes national AI governance (authorities and coordination), sets operational constraints for AI in public administration and the judiciary (human responsibility and authorizations), and introduces criminal/copyright amendments connected to AI use. Organizations should align internal governance to these principles and prepare for delegated implementing legislation and AI Act enforcement via national authorities.
II. Full Description
*Source file:* IT - 20250923 - Legge 23 settembre 2025, n. 132 .pdf
Law 132/2025 sets national principles for AI and delegates multiple areas to future Government decrees, while requiring interpretation consistent with the EU AI Act.
Key provisions visible in the text include:
- Principles: AI research/development/deployment must respect fundamental rights and principles such as transparency, security, data protection, accuracy, non-discrimination, and human oversight.
- Public administration (Art. 14): AI is used to improve efficiency and services, ensuring traceability/knowability and maintaining human responsibility for administrative acts.
- Judiciary (Art. 15): judicial decisions remain reserved to judges; organisational/administrative uses are regulated by the Ministry of Justice; experimentation/use in ordinary courts is authorised by the Ministry of Justice until full AI Act implementation.
- National governance (Arts. 19–20): national AI strategy coordination and designation of AgID and ACN as national AI authorities, with coordination arrangements.
- Delegations (e.g., Art. 16): Government delegated to adopt legislative decrees within 12 months from entry into force on the use of data/algorithms/mathematical methods for training AI systems (within AI Act-covered areas and without additional obligations beyond the AI Act).
- Criminal/copyright amendments: introduces/adjusts offences and penalty enhancements connected to AI use, including a new offence for unlawful dissemination of AI-altered/generated content without consent causing unjust harm.
III. Scope & Application
National framework law setting principles and delegations to the Government on AI, to be interpreted and applied consistently with the EU AI Act. It includes rules on AI use in public administration (traceability/knowability and human responsibility) and on AI in judicial activity (decisions reserved to judges; Ministry of Justice to regulate uses). It designates AgID and the National Cybersecurity Agency (ACN) as national AI authorities and establishes coordination mechanisms; it also amends civil, copyright, and criminal provisions, including a specific offence for unlawful dissemination of AI-generated/altered content without consent causing unjust harm.
IV. Policy Impact Assessment
High impact across public sector, regulated actors, and the legal system: it establishes national AI governance (authorities and coordination), sets operational constraints for AI in public administration and the judiciary (human responsibility and authorizations), and introduces criminal/copyright amendments connected to AI use. Organizations should align internal governance to these principles and prepare for delegated implementing legislation and AI Act enforcement via national authorities.
Primary Focus: national ai governance framework