Italy’s Meloni Target of Deepfake Videos Falsely Showing Anti-Israel Gestures

Italy has suspended a defense cooperation agreement with Israel amid rising tensions between the two countries, a move that has triggered a wave of misinformation in the form of AI-generated videos circulating on social media platforms.

Earlier this month, Italian Prime Minister Giorgia Meloni announced the suspension “in consideration of the current situation.” The decision reportedly came after Israeli forces fired warning shots at an Italian peacekeeping convoy in Lebanon on April 8, escalating diplomatic tensions between Rome and Tel Aviv.

In the wake of this development, two AI-generated videos purporting to show Meloni expressing antagonism toward Israel have gone viral across multiple platforms, misleading viewers about the extent of Italy’s policy shift toward Israel.

The first fabricated video depicts Prime Minister Meloni seemingly ignoring Israeli Prime Minister Benjamin Netanyahu’s handshake attempt during what appears to be a diplomatic meeting. In the manipulated footage, Meloni is shown wearing a scarf resembling the colors of the Palestinian flag, with text overlay stating “Giorgia Meloni ignoring Netanyahu.”

Digital forensic analysis reveals clear indications of synthetic manipulation. Both leaders’ movements appear unnaturally stiff and robotic, lacking the fluid motion characteristic of genuine video footage. When subjected to deepfake detection tools, specifically the Deepfake-o-Meter developed by the University at Buffalo, four separate detection algorithms identified the video as artificially generated.

A second manipulated video shows Meloni apparently losing her temper during an international conference, angrily shouting and throwing papers in what is falsely described as a gesture symbolizing the termination of all agreements with Israel. This video similarly fails to withstand scrutiny when examined closely.

The facial expressions displayed by Meloni in this footage appear unnatural, particularly in the opening segments. More tellingly, the reactions of others present in the conference hall are incongruent with the apparent dramatic display – their body language remains calm and composed despite Meloni’s purported outburst, indicating artificial composition of the scene.

Deepfake detection tools confirmed suspicions about this second video, with three separate detection algorithms identifying it as AI-generated content. A reverse image search revealed that actual footage of Meloni from a 2023 AI safety summit, where she wore the same attire, had likely been manipulated to create the misleading video.

The circulation of these deepfakes highlights growing concerns about AI’s potential to fabricate convincing political content that can mislead the public, particularly during periods of international tension. The technology used to create these videos has become increasingly sophisticated, making detection more challenging without specialized tools.

While Italy has indeed suspended its defense cooperation agreement with Israel, there is no evidence to support claims that Italy has terminated all contracts or diplomatic relations with Israel. The actual diplomatic situation, while tense, remains far less dramatic than the fabricated videos suggest.

The incident underscores the growing challenge facing governments, media organizations, and citizens in distinguishing authentic content from sophisticated AI-generated material. As artificial intelligence capabilities continue to advance, the potential for misrepresentation of political positions and diplomatic actions poses significant risks to public understanding of international affairs.

Experts warn that such manipulated content can exacerbate tensions in already volatile situations and undermine trust in legitimate news sources. The deceptive videos of Prime Minister Meloni serve as a timely reminder of the need for heightened vigilance and improved detection tools in an era where seeing can no longer be equated with believing.