Salvini's Party Fabricated Attack Images with AI to Frame Immigrants as Criminals
What happened
In April 2025, Italy's Lega per Salvini Premier party published dozens of images on its social media accounts showing what it described as immigrants attacking police officers and women. The posts were presented as documentation of real events. Digital forensics experts identified them as AI-generated or AI-manipulated content, fabricated to look like news photography.
The design choices in the images were deliberate. The attackers were depicted as people of color, often shown carrying weapons, with their faces blurred in a style that mimics the editorial practice of obscuring identities in genuine crime reporting. That blurring was not a privacy measure. It was a credibility device, meant to make synthetic imagery read as sourced, verified documentation of actual incidents. Deputy Prime Minister Matteo Salvini's party maintained that its posts reflected real news events, a claim forensics analysis directly contradicted.
Opposition parties filed a formal complaint with Agcom, the Italian communications authority, arguing that the League had not produced isolated offensive posts but had built a sustained campaign portraying immigrants and Arabs as inherently dangerous. The complaint charged the party with systematically constructing a narrative linking specific ethnic and religious groups to criminality, with the stated purpose of inciting hatred, discrimination, and violence against those communities.
The tactic fits a documented pattern. Far-right political movements across Europe have shifted toward synthetic media as the primary raw material for anti-immigration messaging, replacing typed claims with fabricated visual evidence. A generated image that looks like a news photograph carries persuasive weight that a written assertion cannot. Faces blurred in the style of real journalism lend manufactured content the surface features of accountability. The imagery goes out labeled as fact, and nothing in the social media pipeline interrupts it before it reaches audiences.
The Agcom complaint opens a regulatory process, but it cannot reach back and account for what happened before the complaint existed. The images circulated without any indication that they were AI-generated, without provenance data, and without any platform mechanism flagging them as synthetic. A system that logged what produced the content, when, and by whose instruction would have made the fabrication visible at the moment of publication rather than weeks later. Without a provable record of what a system produced and when, political actors can route fabricated imagery through ordinary posting workflows and absorb the regulatory response as an acceptable cost of the campaign.
Reported impact
- Affected parties
- Not publicly disclosed
- Harm type
- Not publicly disclosed
- Scale
- Not publicly disclosed
- Financial impact
- Not publicly disclosed
- Regulatory action
- Not publicly disclosed
Classification
Relevant governance controls
Governance control mapping is not available for this record.
- No controls mapped
Not publicly disclosed
Control mapping is analytical. It does not state that any control would have prevented the incident.
Sources and evidence
This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.