As Manhattan DA Seizes AI Deepfake Sites, Adult Performers are the Hidden Victims

While the bust focuses on 1,200 identified victims — overwhelmingly women, many public figures— adult performers whose bodies were scraped without consent are the casualties.
Industry17th September 2026
(DTS/Nick_Fancher)

The office of Manhattan DA Alvin Bragg took down a dozen AI-generated deepfake porn websites this week under New York’s intimate-image law. And while the bust focuses on 1,200 identified victims — overwhelmingly women, many of them public figures— adult performers (whose bodies were scraped and repurposed without consent or pay) are also casualties. As debates over AI safety, regulation, and its adverse effects dominate headlines, this is the part of the story untold.

Faking It

After a woman found images of herself on an adult website she had never visited, in sexual situations she had never been in, her report to authorities led to the twelve seized domains as well revelations about deepfakes and who’s responsible for them.

“We would not be here if it wasn’t for the courage of a survivor who came forward,” Bragg said in a press announcement.

The twelve sites, with names like AI Celebs and Deepfake Porn, were cited under Bragg’s Cyber Crime Bureau via a New York State Supreme Court warrant citing unlawful dissemination of intimate images — a statute covering AI-altered material as well as real photographs, carrying up to a year in jail as a misdemeanor.

Bragg called it the largest known takedown of AI-generated celebrity deepfake sites, but by his own account, it’s a fraction of what’s out there. Researchers estimate the real number of deepfake porn sites in circulation runs into the millions. No one has been charged, and the office is still working to identify the administrators behind the twelve domains. Bragg said his real interest lies in “the infrastructure of crime” — the hosts, not the individual users.

The Bodies Used

But the infrastructure runs deeper than hosting. It starts with the source material, and that’s where adult performers come in. Every face-swapped image needs a body to swap onto, and increasingly, that body’s movements, proportions, and scenes are lifted from performers’ own work — extracted, altered just enough to evade recognition, and reassembled into something new.

As MIT Technology Review documented in May 2026, this practice has been constant for years and has only gotten harder to fight as the technology improves. Performers’ work has effectively become training data: their bodies shape how synthetic bodies look and move, so the resulting content competes with the original performer’s income while remaining almost untraceable back to them.

Why Prosecutors Are Targeting Hosting Sites, Not AI Tools

There are three places enforcement can aim in a deepfake case: the person who made the image, the site that hosted it, or the tool that generated it.

The first is nearly useless at scale — the people prompting these images are anonymous and numerous. The third is where the more consequential fight is happening. Minnesota’s HF 1606, effective August 1, 2026, is the first U.S. law to regulate AI nudification tools directly, with penalties up to $500,000 per violation. xAI sued the state in July to block it.

Bragg, by contrast, is working with a misdemeanor — a workable tool against anonymous domain owners. And notably absent from any of these enforcement conversations are the performers whose bodies fuel the underlying models. No statute names them; and the takedown numbers don’t count them.

The Grok Deepfake Lawsuits

While the Manhattan DA built a case against a dozen no-name sites, 2026’s biggest non-consensual AI imagery story has been about Grok, xAI’s chatbot. After xAI opened image editing on X last December, users discovered they could strip clothing from photographs of real people, and did so at volume. The company’s response was to restrict the feature to paying subscribers — limiting who could access it, not stopping the behavior itself.

Litigation followed anyway. Ashley St. Clair (the mother of Elon Musk‘s child, in the news for revelations about her ex) sued xAI in January over sexually explicit images the chatbot generated of her. California Attorney General Rob Bonta sent a cease-and-desist around the same time. In March, a class action in the Northern District of California alleged Grok had produced child sexual abuse material from real photographs of minors. Regulators in the EU, UK, and Canada have since opened inquiries. xAI maintains it strictly prohibits generating nude or sexualized images without consent, and the allegations remain unproven.

Critics highlight the disparity: nobody has seized anything from xAI, likely because of who the defendant is. And if named public figures struggle to get traction against a company with resources, adult performers — an already stigmatized group with far less institutional sympathy — have essentially no path to the same kind of attention or legal response.

Take It Down Act, DEFIANCE Act, and NO FAKES Act

The current federal takedown landscape includes a few different efforts, none built with performers specifically in mind. The Take It Down Act, enforced by the FTC since May 19, 2026, requires covered platforms to remove reported non-consensual intimate images within 48 hours. That’s different from what Bragg did this week: the FTC’s rule takes content down, while Bragg targeted the businesses. Neither compensates victims.

The DEFIANCE Act does compensate — a federal civil claim with damages up to $250,000. It passed the Senate unanimously in January 2026 and has stalled in the House since.

Then there’s the NO FAKES Act, which would establish a federal right over one’s own voice and likeness. It cleared Senate Judiciary in June and has similarly gone nowhere. Forty-five states now have some version of a deepfake law, but the federal law that would let a victim sue has yet to pass.

The Porn Paradox

The same enforcement machinery built to protect people from non-consensual imagery has also been used to strip consensual, lawfully produced adult content. Broad takedown tools, blunt content moderation, and platform overcorrection don’t reliably distinguish between a stolen deepfake and a performer’s own legal work. The people most exposed to synthetic exploitation are frequently the same people penalized by the systems meant to stop it.

That’s the paradox sitting underneath Bragg’s twelve seized domains: adult performers are both the uncredited raw material for a growing industry of non-consensual imagery, and among the first casualties when platforms try to clean it up. They’re not named in the 1,200-victim count. They don’t have a cease-and-desist or a class action with their name on it. They’re simply the ones whose work made the technology possible — and whose consent nobody thought to ask for either way.

About the author
Wet Ink Team's avatar
Wet Ink Team

Related articles

Join the Wet Ink newsletter

By clicking “Subscribe” button you agree to our Terms of Service

Discover more from Wet Ink Magazine

Subscribe now to keep reading and get access to the full archive.

Continue reading