Opposition in Italy Files Complaint Regarding Far-Right Deputy Prime Minister’s Use of ‘Racist’ AI Images
Controversy Over AI-Generated Images in Italian Politics
In Italy, opposition political groups have voiced concerns regarding the use of AI-generated images by the far-right party led by Deputy Prime Minister Matteo Salvini. They have lodged a complaint with the Italian communications watchdog, Agcom, arguing that these images are deeply problematic, describing them as “racist, Islamophobic, and xenophobic.”
Allegations Against the League Party
The centre-left Democratic Party (PD), along with the Greens and Left Alliance, submitted their complaint on a Thursday, claiming that the League’s social media posts exhibit nearly all categories of hate speech. These AI-generated images have surfaced on multiple platforms, including Facebook, Instagram, and X (formerly Twitter), and often depict men of color, commonly armed, in aggressive scenarios targeting women and police officers.
Senator Antonio Nicita from the PD emphasized the severity of the situation, pointing out that the images depict immigrants and Arabs in a negative light, labeling them as potential criminals. “These images not only convey violence but are also misleading,” Nicita stated. The faces of those attacked are blurred, which he suggests is an intentional act to imply that such violent incidents are commonplace, despite lacking verifiable details.
Party’s Response
A spokesperson for Salvini’s party confirmed that “some of the pictures” shared online were indeed “generated digitally.” They defended the use of these images by asserting that every post has a basis in real events reported by Italian media, including specific names, dates, and locations. The party contended that if the true nature of these events seems harsh, the blame should not fall on the reports but rather on the realities being depicted.
Concerns from Experts
Salvatore Romano, the head of AI Forensics, expressed that the AI images used by the League have distinct characteristics of artificial intelligence, such as being out-of-context and overly focused. He noted that as AI technology improves, these images are becoming alarmingly realistic, raising concerns about their potential to misinform the public.
The complaint to Agcom highlights specific incidents where these AI-generated images were used alongside reputable media reports. For instance, one League post claims “a foreigner attacks the train conductor,” presenting a digitally altered image of a suspect without context or confirmation of their nationality from the original news article. Similarly, another image depicted a family appearing to berate a young girl, feeding into stereotypes without any factual basis from the news report cited.
Growing Popularity of AI-Generated Content
The deployment of AI-generated imagery for political propaganda by far-right parties has become increasingly common. This trend began gaining momentum during the European elections, as parties began using manipulative images aimed at stoking fears surrounding immigration and demonizing political figures. Discussion has indicated that platforms like Twitter, particularly during the American election cycle, have normalized such practices further.
Romano pointed out that while social media platforms have an obligation to mitigate the risks associated with AI-generated content by labeling it clearly, these measures often fall short in practice.
Implications for Regulation
Should Agcom find the content in question offensive, it holds the authority to enforce penalties under the EU’s Digital Services Act, which could include removing posts, banning accounts, and imposing fines on social media platforms for non-compliance. Recently, Agcom imposed a fine of €5.85 million on Meta for violating advertising regulations.
As this situation evolves, public discourse continues about the ethical implications of using AI technology in political messaging and the responsibilities of both political parties and social media platforms to ensure accurate representation and the responsible use of technology in informing voters.
