Shocking pretty much no one, a new report claims that AI-generated clear images of Taylor Swift came from the Internet's most notorious cesspool: 4chan.
In the findings shared with Rolling rock, research firm Graphika said it supplied the images to a particular message board community on 4chan that was essentially making a game of prompting AI image generators to include security measures and create graphic images of famous people. (The technology publication 404 He similarly traced the images back to 4chan in a story published last month.)
And while the fake images of Swift garnered the most attention, investigators said she wasn't the only target, nor the most important of the message board's denizens.
“While Taylor Swift's viral pornographic photos have drawn attention to the issue of non-consensual AI-generated images, she is not the only victim,” Graphika senior analyst Cristina López G. said in a statement. “In the 4chan community where these images originated, he is not even the most frequently targeted public figure. This shows that anyone can be targeted in this way, from global celebrities to school children.”
On the message board, per The New York Times, users searched for and shared “tips and tricks to find new ways to bypass filters.” Particularly realistic images received upvotes, while “mistakes” – such as a prompt that resulted in an AI-generated image of a celebrity in a swimsuit, rather than nude – were lamented. (Similar “games” have been used for other purposes on 4chan, as well as on platforms such as Telegram and WhatsApp, including extensive efforts to detect alleged voter fraud.)
López G. added: “These images came from a community of people motivated by the 'challenge' of bypassing the safeguards of productive AI products, and the new restrictions are seen as just another obstacle to 'defeat'.” It is important to understand the gamified nature of this malicious activity in order to prevent further abuse at the source.”
The fake, clear images of Swift first appeared on 4chan on January 6, but just a week later they started going viral on more mainstream platforms like Telegram and X (formerly known as Twitter). The proliferation of images was so bad that X had to block searches for “Taylor Swift” for several days while he removed the images. Swift's outraged fans, who helped remove many of the fake images through mass reporting, also called for more protection, however, as Rolling rock mentioned earlier, these will not come easily.
SAG-AFTRA then condemned the images as “disturbing, harmful and deeply disturbing” and said “the development and dissemination of false images … without someone's consent should be made illegal.” White House press secretary Karine Jean-Pierre even addressed the issue, calling the images “disturbing,” demanding that social media companies take steps to more forcefully enforce their content moderation policies, and calling on Congress to pass protective legislation.
from our partners at https://www.rollingstone.com/music/music-news/taylor-swift-explicit-ai-images-4chan-game-1234961414/
