Following the enactment of the EU AI Act, the first AI regulation globally; now Italy has taken a major step towards regulation of the Artificial Intelligence technology. On 17th September, 2025 , the parliament has approved the Bill no. 1146, a comprehensive law regulating artificial intelligence, marking a decisive step in Europe’s attempt to balance innovation with public safeguards. Bill No. 1146 – “Provisions and Delegations to the Government on Artificial Intelligence” sets out a wide-ranging framework to govern AI use, with a particular focus on deepfakes, child access to AI systems, and data privacy protections. The law, spearheaded by Prime Minister Giorgia Meloni’s government, aligns in many respects with the EU’s own forthcoming AI Act, but goes beyond in certain specifics.
The move positions Italy ahead of the EU’s own bloc-wide AI Act, which is expected to come into effect in 2026. While the EU legislation aims to harmonize rules across member states, Italy’s government has chosen to act swiftly, introducing stricter national provisions that reflect growing public concern over the unchecked spread of generative AI technologies.
A First-of-Its-Kind National Framework
Bill No. 1146 was debated and passed in the Italian Senate before receiving final approval from the lower house of Parliament this month. The legislation sets out delegations to the government to issue secondary rules and guidance in the months ahead, but already provides clear principles for AI governance.
According to the text, the law seeks to ensure that AI systems are developed and deployed in line with human dignity, safety, privacy, and democratic values. The bill is structured into six chapters. Its contents include both high-level principles and detailed sectoral rules. Below are its core elements:
| Area | Provisions in Bill 1146 |
| Principles and Foundational Norms | Chapter I sets out guiding principles: respect for fundamental rights (both under the Italian Constitution and the EU), fairness, reliability, safety, quality, proportionality, human autonomy, prevention of harm, intelligibility, transparency, explainability, equality, protection of democratic processes, and the accessibility of AI systems to persons with disabilities. |
| Media, Expression, and Minors | The law explicitly protects freedom and pluralism of the media, freedom of expression, fairness of information. It places specific limitations on AI use by minors. |
| Sectoral Application | In Chapter II, Bill 1146 provides tailored rules for: • Healthcare (Art. 7): AI systems must support professionals, not replace medical decision-making. Patients must be informed about AI use and algorithmic logic. Discrimination or conditioning of healthcare access is prohibited. • Scientific research and experimentation involving sensitive data (Art. 8) under ethics committee oversight. • Personal data processing, employment, public administration, justice, etc. |
| Oversight & Authorities | Chapter III establishes that the national AI strategy is to be prepared by the Presidency of the Council of Ministers, in charge of technological innovation and digital transition, with the involvement of the National AI Authorities. It designates Agenzia per l’Italia Digitale (AgID) and Agenzia per la Cybersecurity Nazionale (ACN) as the bodies that will implement national and European AI regulation. |
| Criminal and Intellectual Property Law Changes | Chapter IV amends copyright law: works created with AI may be protected if they result from human intellectual effort. Chapter V introduces Article 612-quater into the Criminal Code, criminalizing the dissemination of deepfakes that cause unjust harm to the person depicted. |
| Delegation and Alignment with EU Law | The bill delegates the Italian Government to adopt legislative decrees within twelve months after the law enters into force to adapt national legislation fully to the EU AI Act. Also defines rules for unlawful AI system development and use. |
GOVERNMENT’S POSITION
The legislation was championed by Prime Minister Giorgia Meloni’s government, which framed the bill as both a protective measure for citizens and a signal of Italy’s ambition to be a European leader in AI governance. While Meloni herself did not issue a sweeping public statement during the vote, ministers emphasized the importance of creating a framework that prevents harm while enabling innovation.
Reuters highlighted that the law reflects mounting concern in Italy over the societal risks of generative AI, particularly the surge of AI-driven disinformation and deepfake scandals that have dominated headlines in recent months. Lawmakers argued that without a clear regulatory framework, the spread of synthetic media could undermine public trust in institutions, elections, and journalism.
Looking Ahead
The passage of Bill No. 1146 does not end the debate, rather, it opens a new chapter. Much of the law’s impact will depend on how the government exercises its delegated powers in drafting detailed regulations. Observers expect contentious discussions around implementation timelines, enforcement mechanisms, and alignment with the EU AI Act.
Still, Italy’s decision underscores the political urgency of addressing AI risks. As deepfakes multiply, disinformation campaigns evolve, and children’s digital environments grow more complex, Italy’s lawmakers have chosen to move quickly rather than wait for Brussels.
Whether this bold step proves to be a model for the EU or an overreach that burdens Italian industry will become clear in the coming years. For now, what is certain is that Italy has written itself into the history books as the first EU nation with a comprehensive AI law , a law that could shape Europe’s digital future well beyond its borders.