Italy Becomes the First in Europe to Set National AI Rules
On September 17, 2025, the Italian Parliament approved a groundbreaking law on artificial intelligence (Law no. 132/2025, the “AI Law”), marking a significant milestone as the first of its kind in the European Union. The AI Law establishes core principles governing AI use, delineating responsibilities for AI governance and supervision in Italy. It provides specific rules for critical sectors including healthcare, public administration, justice, labor, intellectual property, and criminal law. This law will come into force on October 10, 2025, and aims to enhance the existing framework established under Regulation 2024/1689 (“AI Act”) without introducing new compliance obligations.
Principles Underpinning the AI Law
The AI Law is underpinned by several key principles:
Human-Centered Vision of AI:
AI is intended to enhance human decision-making rather than replace it. AI systems must comply with fundamental rights and EU law, emphasizing values such as transparency, fairness, safety, privacy, gender equality, and sustainability. Human oversight is mandated, ensuring that individuals can understand, monitor, and intervene throughout the AI system lifecycle.
Protection of Personal Data:
AI systems must handle data lawfully and transparently, ensuring users are informed in clear, accessible language about data usage and associated risks. This principle protects freedom of expression and media pluralism, preventing AI from compromising the objectivity and fairness of information.
Protection of Democratic Debate:
The law prohibits AI applications that could interfere with democratic institutions or distort public debate, addressing concerns regarding AI’s potential to spread disinformation and manipulate opinions.
AI as a Key Lever for Economic Growth:
Italian public authorities support AI development to enhance human-machine interaction across various sectors. They promote AI adoption to boost productivity and foster the growth of new businesses, particularly focusing on strengthening Italy’s industrial base, comprising many micro, small, and medium-sized enterprises.
Sectoral Provisions
The AI Law outlines specific provisions for various sectors:
Healthcare:
AI systems are designed to support medical professionals, assisting with prevention, diagnosis, and treatment under strict supervision. Transparency is emphasized, as patients must be informed when AI technologies are part of their care.
Labor Sector:
AI should enhance working conditions while safeguarding employee well-being. Workers must be informed about AI usage in recruitment and daily operations, and AI applications must respect fundamental rights, prohibiting discrimination based on various personal characteristics.
Public Sector:
Public administrations can utilize AI to improve efficiency and service quality, but AI serves solely as a support tool, with human officials retaining full responsibility for decisions and actions.
Justice System:
AI can aid in organizing judicial services and easing court workloads, but all legal decisions remain the exclusive domain of judges, with training provided to enhance their understanding of AI risks and benefits.
Supervisory and Enforcement Authorities
The AI Law establishes oversight bodies at the national level, aligning with the EU AI Act:
- The Presidency of the Council oversees the national AI strategy, approved bi-annually.
- The Coordination Committee promotes research and adoption of AI systems.
- The Agency for Digital Italy (AgID) acts as the notifying authority, ensuring compliance with national and EU legislation.
- The National Cybersecurity Agency (ACN) oversees AI security and acts as the point of contact with the EU.
Changes to Criminal and IP Law
The AI Law introduces significant changes to criminal law:
- A new offense for the unlawful dissemination of AI-manipulated content (deepfakes) may result in imprisonment.
- Aggravating circumstances are introduced for crimes committed using AI tools, leading to stricter penalties.
- Unauthorized text-and-data mining (TDM) is criminalized, while lawful TDM for training AI models is permitted under specific conditions.
In terms of copyright, the law updates the Italian Copyright Law to clarify protections for “works of human authorship” and restricts AI-generated content without human creative input.
Investment in AI and Emerging Technologies
To foster development in AI and cybersecurity, the Italian government is allocating up to €1 billion in public investment targeted at innovative small and medium-sized enterprises in these fields.
Implications for Companies
As Italy implements these new regulations, companies must adapt to the evolving legal landscape. Organizations are encouraged to update their management and control models to mitigate AI-related risks and align with the AI Law’s requirements. Monitoring the implementation of associated decrees will be critical to address liability issues and ensure compliance.
The AI Law sets a precedent in Europe, establishing a framework that balances innovation with ethical considerations and regulatory oversight, signaling Italy’s commitment to responsible AI development.
