AI deepfakes of Katy Perry and Rihanna supposedly attending the 2024 Met Gala are going viral on social media in the latest incident of generative AI becoming more dangerously deceptive.
Images of the celebrities first trended on X (formerly Twitter) and Instagram last Monday while the actual event at the Metropolitan Museum of Art is starting.
(Photo : Olympia de Maismont/AFP via Getty Images)
Although the images were quickly flagged by several keen-eyed fans to be AI-generated, the pictures already made their way to the hundreds of thousands of fans monitoring the event live.
It did not help that bots boosted the posts to many people's feeds, further spreading the posts' reach to unaware users.
Perry later shared that even her mother was fooled by the AI-generated images of her. The singer clarified that she had never attended this year's Met Gala because she "had to work."
People Magazine earlier reported that Rihanna was unable to attend the event due to flu.
Also Read: AI Deepfakes of Piers Morgan, Oprah Used in Fake Online Ads
AI Deepfakes Become More Deceptive
While the singer-celebrities did not seem to mind the viral AI images, the rise of these near-realistic deepfakes depicts the worrying trend of generative AI's impact on information democracy.
Concerns for AI deepfakes have been growing since the explicit deepfake images of Taylor Swift first went viral on X earlier this year.
Since then, more celebrities' likenesses have been noted as being used in AI-powered scams and fake advertisements. The trend has even spread to online influencers, the Financial Times reported.
It is another issue if the images were produced by the more popular AI image generators following the companies' promises to prevent the digital replication of notable personalities' likenesses.
It remains unclear which AI image generator the deepfakes came from.
Social Platforms Crack Down on AI Deepfakes
With most AI-powered disinformation coming from their platforms, social media companies like X, Facebook, Instagram, YouTube, and TikTok have started rolling out labels to help users identify AI-generated media.
The policies are expected to complement a proposed bill in Congress to require proper labels for all AI-generated media in anticipation of its surge ahead of the election period.
Related Article: New Bipartisan Bill Seeks to Require AI-Generated Media to be Properly Labeled
