New anti-deep fake law: operative offence with aggravating circumstance for market rigging
The new legislation punishes the spread of deep fakes with AI and introduces stricter penalties for financial crimes with the use of artificial intelligence.
Key points
As of today the Artificial Intelligence Act is in force. And also operational from today is theoffence of deep fakes, a common aggravating circumstance and three special ones to raise a sort of criminal shield against the illicit use of algorithms. Central to this is the introduction of a new offence to punish the publication and dissemination of images, videos or voices falsified or altered through artificial intelligence systems, capable of misleading people as to their authenticity and causing harm to the person concerned.
The offence is punishable on complaint, but is prosecuted ex officio when the act is connected with another offence that is itself prosecutable ex officio and or if the act has been committed against a person who is incapacitated or a public authority by reason of the functions exercised. The conduct must be such as to cause a damage that, in the absence of a specific indication as to the need for profit, may also be of a non-pecuniary nature.
Against deep fakes
The aim of the regulation is to combat the spread of deep fakes, i.e. "an image or audio or video content generated or manipulated by artificial intelligence that resembles an existing person, object, place, or other entity or event and that would appear falsely authentic or truthful to a person", as defined by the AI Act.
The new offence fills a protection gap in our criminal justice system with regard to the dissemination of deep fake porn with victims who are over 18 years old, conduct that until now has not been punishable on the basis of Article 612-ter of the Criminal Code, unlike virtual images of minors.
Space is then given to a common aggravating circumstance to make the criminal sanction heavier when the act was committed through the use of artificial intelligence systems, as the same systems can be classified as an 'insidious means, when the consequences of the offence have been aggravated or when the use of Ai programmes has hindered the defence, public or private. It is up to the judicial authorities to assess, case by case, possibly with expertise, when the use of algorithms has contributed to surprise the attention of the offended person.


