Who Is Behind the Spread of Taylor Swift’s Deepfake Nudes?

The fallout from last week’s posting of nonconsensual artificial intelligence-generated deepfake nude imagery of pop superstar Taylor Swift has not slowed, with members of Congress speaking out on the matter, X (née Twitter) reinstating the ability to search Swift’s name after it had been disabled as the images tore across the platform, actors union SAG-AFTRA condemning the false images and legions of Swifties retaliating online.

Then came the burning questions: What is the source of these images and how did they spread?

More from The Hollywood Reporter

Starting on Wednesday, one of the most proliferated fake images of Swift to initially appear on X attracted over 45 million views, 24,000 reposts, and hundreds of thousands of likes and bookmarks. It took 17 hours for the now Elon Musk-led microblogging platform to suspend the account of the verified (or paid, under Musk’s changes) user for violating policy.

The false images in question feature Swift in suggestive positions at Kansas City Chiefs games, which the star has been attending of late in support of her boyfriend, Travis Kelce.

According to reporting from 404 Media‘s Emanuel Maiberg and Sam Cole, the fake images jumped from 4chan — the Wild West website heavily associated with the Q-Anon conspiracy theorists — and a group on the Telegram app where users share AI-generated images of celebrity women. AI renderings of Swift also appeared on the website Celebrity Jihad, which is known for posting leaked private videos and imagery — and deepfakes — of celebrities; this is said to be done as satire.

404 reports that the images of Swift do not appear to be a result of superimposing Swift’s face over a pornographic image but were instead created with a commercially available AI image generation tool. That tool appears to have been Microsoft’s AI image generator called Designer; users in the Telegram group and on 4chan also share prompts to work around Microsoft’s safeguards preventing the generation of explicit imagery.

“We are committed to providing a safe and respectful experience for everyone,” a Microsoft spokesperson told The Hollywood Reporter via email Thursday, in response to reports of Designer being used to create the controversial images. “We are continuing to investigate these images and have strengthened our existing safety systems to further prevent our services from being misused to help generate images like them.”

In the firestorm that ensued as the fake Swift images spread online, at least one account that seemingly brought them from these dark corners of the web to X was the username @Zvbear, which was set to private Thursday. It reportedly then had around 83,000 followers, but has shed several thousand since.

Other accounts associated with the user on X and TikTok have been shut down, following a report by Newsweek that quoted multiple X posts from @Zvbear before the account went private, including, “My Taylor post went viral and now everyone is posting it.”

Lawmakers are beginning to introduce legislation to prevent this type of abuse. Rep. Joe Morelle (D-NY) renewed efforts last week to make the nonconsensual sharing of digitally altered explicit images a federal crime that includes jail sentences.

The White House has also commented on the matter, with Press Secretary Karine Jean-Pierre taking social media platforms to task for their role in spreading the imagery.

“Too often, we know that lax enforcement disproportionately impacts women,” Jean-Pierre said. “While social media companies make their own independent decisions about content management, we believe they have an important role to play in enforcing … their own rules to prevent the spread of misinformation and nonconsensual, intimate imagery of real people,” she said.

Swift and her team have remained publicly silent on this controversy so far, and did not respond to multiple comment requests from THR.

But one source hinted at legal action against the perpetrators while speaking with the Daily Mail.  “Whether or not legal action will be taken is being decided, but there is one thing that is clear: These fake, AI-generated images are abusive, offensive, exploitative and done without Taylor’s consent and/or knowledge,” said the anonymous source.

Best of The Hollywood Reporter