Artificial intelligence is no longer merely a technological or economic matter: it has become, in every respect, a legal and criminal law issue. With Law no. 132/2025, the Italian legislature has taken a significant first step in this direction, introducing into the Criminal Code the new Art. 612-quater, which sanctions the unlawful dissemination of content generated or manipulated by means of artificial intelligence systems — the so-called deepfakes.
This provision is destined to have a concrete impact on the lives of many people: from victims of false and defamatory content circulated online, to companies managing reputational crises triggered by fabricated images or videos, to legal professionals who must navigate a regulatory framework that is still evolving. In this article, we aim to clarify what the new provision entails and what its practical implications are.
What Is a Deepfake and Why Does Criminal Law Address It
The term deepfake refers to content — video, images, audio, or text — created or convincingly altered through artificial intelligence algorithms, to the point of appearing entirely credible to those who receive it. The unlawful applications are numerous: from the defamation of public figures to the creation of non-consensual sexually explicit content, from the manipulation of information to commercial fraud.
Prior to the entry into force of Law no. 132/2025, the Italian criminal justice system lacked a dedicated instrument to address this phenomenon. Judges were compelled to resort to general offences — such as defamation, privacy law provisions, or those governing the unlawful processing of personal data — with evident difficulties in application. The new Art. 612-quater fills, at least in part, this legislative gap.
The Structure of the New Offence: Essential Elements
An analysis of the new provision reveals a composite offence, the fundamental elements of which are worth examining in detail:
- The actus reus concerns the dissemination of content generated or altered by means of AI systems. Merely creating the material is insufficient: it is its circulation — through social media platforms, messaging services, websites, or any other channel — that constitutes the offence.
- The requirement of unjust harm constitutes a central element of the offence. The provision requires that the conduct result in unjust harm to the person depicted or involved in the artificially generated content. This element limits the scope of application, in principle excluding satirical or artistic uses that cause no concrete injury.
- The procedural aspects represent a sensitive issue on which expert debate remains ongoing: the choice between prosecution upon complaint by the victim (procedibilità a querela) or ex officio prosecution (procedibilità d'ufficio) has a profound bearing on
