Winter Olympics Targeted by Wave of AI-Generated Misinformation

The 2024 Winter Olympics in Milan has become a hotbed for digitally manipulated content, as false videos and AI-generated disinformation flood social media platforms, according to an investigation by Euronews’ fact-checking team, The Cube.

Among the most prominent examples is a fabricated video purportedly showing U.S. Vice President JD Vance being booed during the Olympic opening ceremony. The clip, which garnered over 4 million views on X (formerly Twitter), depicts Vance standing in the crowd with a concerned expression as loud jeers erupt in the background.

Fact-checkers have confirmed the video is not authentic. While Vance did attend both the opening ceremony and a women’s ice hockey preliminary round, the viral clip contains multiple inconsistencies. Official photographs show Vance wearing a formal suit at the ceremony, not the windbreaker seen in the manipulated video. Additionally, experts at Sensity AI identified telltale signs of AI manipulation, including unnatural eye movements, briefly disappearing pupils, and a section where the video plays in reverse while audio continues uninterrupted.

Sensity AI demonstrated how easily such content can be created, using commercially available AI tools to generate a similar deepfake from a still photograph for approximately €12.

Though the viral video is fake, legitimate reports confirm that Vance did face boos from the crowd when images of him and his wife appeared on screen after Team USA passed during the opening ceremony. CBC News commentators and journalists attending the event documented this reaction.

Russian disinformation campaigns have also leveraged the Olympics to target Ukrainian refugees and athletes. Multiple fabricated videos circulating on Telegram falsely claim to show Ukrainian refugees protesting outside Italian embassies, demanding the cancellation of the Olympic Games on the grounds that they represent an “extravagant expense” while Italy “ignores the needs of Ukraine.”

In reality, no credible media outlets have reported such protests by Ukrainian refugees. The actual demonstrations that occurred were organized by anti-ICE groups, environmental activists concerned about the Olympics’ impact, and pro-Palestinian organizations.

Another sophisticated piece of misinformation features a doctored CBC news report with an AI-generated voiceover, falsely claiming that Ukrainian athletes are being segregated at the Games due to aggressive behavior. Fact-checkers identified the footage as manipulated stock material that bears no resemblance to the original CBC report by journalist Adrienne Arsenault, which focused on the Olympic village without any reference to Ukrainian refugees or segregation.

Ukraine’s Centre for Countering Disinformation has documented additional false narratives targeting the country’s Olympic team, including claims that anti-doping controls were relaxed for Ukrainian athletes and that translators abandoned Ukrainian delegations. The organization described these as “coordinated” attempts by Russian networks to discredit Ukrainian athletes at the Games, which Russia is banned from participating in.

In a separate incident, social media users spread a fabricated story about a nonexistent “Princess Frida” of Sweden, who allegedly criticized former U.S. President Donald Trump and invited Canada to join the European Union. The posts, which reached thousands of viewers and sparked genuine debate about Canada’s potential EU membership, used an image of Swedish cross-country skier Frida Karlsson, who won gold in the women’s 10km + 10km skiathlon.

Sweden’s royal family roster confirms no princess by that name exists, and Karlsson has made no such political statements.

The proliferation of these sophisticated fakes highlights growing concerns about the potential for AI-generated content to spread rapidly across social platforms during high-profile global events, potentially influencing public perception and political discourse.