The Italian Government has approved the first two draft legislative decrees implementing Law No. 132/2025 on artificial intelligence as part of the national implementation of the framework established by the AI Act.

The measures address the national AI governance framework, AI literacy and training initiatives, the use of AI systems within public administration and law enforcement activities, and certain civil liability aspects related to the deployment of such technologies.

The first draft decree designates the Italian Digital Agency (“AgID”) as the national notifying authority and the National Cybersecurity Agency (“ACN”) as the market surveillance authority and single national contact point for the purposes of the AI Act.

The draft decree places significant emphasis on AI literacy measures, it provides for the integration of artificial intelligence into educational programs, the update of STEM guidelines and national curricula, and the promotion of the conscious, critical and responsible use of generative AI tools.

The decree allocates an additional EUR 100 million under the National Program “Schools and Skills 2021-2027” to support teacher training relating to the use of social media, digital platforms and AI systems. These resources are in addition to the EUR 100 million already allocated under Ministerial Decree No. 219/2025 and the public call issued on 27 March 2026 concerning local training hubs on the use of AI in schools, bringing the total amount dedicated to training school personnel and promoting AI-related skills in schools to EUR 200 million.

AI literacy measures are also extended to Provincial Centers for Adult Education (“CPIA”), public administrations, regulated and non-regulated professions and the healthcare sector. This approach is aligned with the AI literacy obligation set out in Article 4 of the AI Act, which requires providers and deployers of AI systems to ensure a sufficient level of competence among individuals involved in the use of such systems.

With regard to employment relationships, the draft decree provides that decisions supported by AI systems must ultimately remain under the responsibility of a natural person vested with effective and autonomous decision-making powers. Employees are entitled to receive an intelligible explanation of decisions affecting them, including information on the role played by the AI system and the main parameters considered. Any dismissal adopted in breach of the prohibition on fully automated decision-making is deemed null and void.

The draft decree also requires AI systems capable of affecting work organization or safety-related decision-making processes to be considered within the risk assessment framework set out in Legislative Decree No. 81/2008.

In addition, the decree establishes the Italian AI Regulatory Sandbox pursuant to Article 57 of the AI Act, aimed at supporting innovation, facilitating access to the European Union market, and promoting cooperation among businesses, universities, and research institutes.

As regards enforcement, the measure mirrors the sanctions framework established by the AI Act, providing for fines of up to EUR 35 million or 7% of annual worldwide turnover for the most serious infringements. The draft decree also extends trade secret protection to data, algorithms and mathematical methods used to train AI systems, subject to the applicable legal requirements.

The second draft decree governs the use of AI systems in law enforcement activities and addresses civil and criminal liability issues. It introduces procedural safeguards for biometric systems and facial recognition technologies, including human oversight requirements, fundamental rights impact assessments, logging obligations and restrictions on the use of biometric databases created through untargeted scraping or in breach of data protection rules.

The decree also introduces mechanisms aimed at reducing information asymmetries in civil litigation by enabling courts to order the disclosure of evidence relating to the functioning of AI systems. It further establishes a rebuttable presumption of causation where damage results from a breach of obligations under the AI Act and allows injured parties to bring direct claims against the insurer of the liable entity.

The practical implementation of the measures set out in the draft decrees will be relevant to the operation of the national framework.