This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Italian Prime Minister Giorgia Meloni initiated a defamation lawsuit against a father and son who created and uploaded non-consensual deepfake pornographic videos of her to a U.S.-based website in 2020. The videos were viewed millions of times, and Meloni is seeking €100,000 in damages, which she has pledged to donate to a fund for victims of domestic violence. The case has become a landmark effort in Italy to address the misuse of AI-generated synthetic media and to encourage other victims of digital abuse to pursue legal action.
In 2020, alleged deepfake pornographic videos falsely depicting Italian Prime Minister Giorgia Meloni were uploaded to a U.S. website and viewed millions of times. At the time, she was leader of the Brothers of Italy but not yet PM. In 2024, now serving as PM, Meloni reportedly sued Alessio Scurosu, 40, and his father, Roberto Scurosu, 73, seeking €100,000 in damages. She testified in court, calling the act "a form of violence."
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to gain a personal advantage over others such as through cheating, fraud, scams, blackmail or targeted manipulation of beliefs or behavior. Examples include AI-facilitated plagiarism for research or education, impersonating a trusted or fake individual for illegitimate financial benefit, or creating humiliating or sexual imagery.
Also classified as
The non-consensual pornographic deepfakes were uploaded to a public website where they were viewed millions of times, exposing the public to toxic and non-consensual sexual content.
The perpetrators used AI to create humiliating and sexually explicit imagery of a specific victim to degrade and defame her, which aligns with targeted manipulation and harassment.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was caused by the intentional actions of human perpetrators who chose to create and upload defamatory deepfake videos using deployed AI tools.
Added categories peak at 2: Minor (Psychological Harm, Direct)
Two Italian citizens created and uploaded non-consensual pornographic deepfakes of politician Giorgia Meloni in 2020. While the incident represents a severe violation of personal privacy and targeted harassment of a political figure, its national security impact remains minor as it was carried out by private individuals without broader coordination or intent to disrupt state functions.
The reports describe the perpetrators using open-source face-swapping technology to overlay Meloni's face onto another person's body. The AI functioned as a tool to perform this specific, well-defined image manipulation task based on direct human inputs.
Risk Level 3: Limited Risk. The EU AI Act classifies AI-generated content such as deepfakes under limited risk, requiring transparency obligations to ensure users are informed of the AI's nature.
The AI technology was explicitly used to generate deepfake pornographic videos and images by superimposing the victim's face onto explicit content.
The reports provide consistent and clear details about the incident, including the identity of the victim, the perpetrators, the legal proceedings, the platform where the videos were uploaded, and the nature of the AI technology used. There is very little ambiguity regarding the core facts of the case.
None
Technology
Accessible Face-Swapping AI
User-friendly open-source software made deepfake creation easy.
Realistic Deep Learning Models
Advanced AI algorithms generated highly convincing synthetic media.
Data Inputs
Unconsented Facial Images
Publicly available photos of Meloni were harvested without her consent.
Explicit Target Videos
Existing pornographic videos were used as templates for face swapping.
Human Factors
Intent to Humiliate and Abuse
Perpetrators aimed to degrade and punish the victim through explicit media.
Public Consumption Appetite
Millions of online views driven by user demand for explicit content.
Process and Methods
Foreign Website Hosting
Uploading to a US-based site bypassed immediate European intervention.
Lack of Proactive Filtering
Hosting platforms lacked automated checks to block non-consensual uploads.
Management
Poor Platform Content Moderation
The adult website failed to moderate and remove non-consensual content.
Flawed Search Engine Ranking
Search algorithms prioritized explicit deepfakes over news articles.
Regulatory Environment
Absence of Deepfake Laws
No specific laws criminalized non-consensual deepfakes in 2020.
Cross-Border Legal Gaps
Jurisdictional issues hindered fast removal from foreign-hosted sites.
AI system involved
An AI system was involved, but the reports do not identify it.
Identified by automated extraction; can differ from the alleged developers above.
