The Manhattan District Attorney’s Office just executed the largest coordinated legal action against deepfake pornography sites, seizing 12 websites that collectively targeted roughly 1,200 victims. It’s the kind of enforcement action that deepfake victims and their advocates have been demanding for years.

The timing isn’t coincidental. The seizures come as new analysis reveals the scope of the problem extends far beyond celebrities. Researchers tracking 160 deepfake pornography websites found images of politicians from 22 European countries. Nearly all of them are women.

What the seizures actually mean

The Manhattan DA’s action is legally significant for a few reasons. First, it’s a seizure, not just a cease-and-desist letter. The sites are gone, at least for now. That’s a stronger remedy than most victims have been able to secure through civil litigation alone.

Second, it targets the infrastructure itself. Going after the websites, rather than just individual users or creators, disrupts the distribution channel. It won’t stop deepfake creation, but it makes it harder to profit from it at scale.

The challenge, as always, is jurisdiction. Deepfake sites pop up faster than prosecutors can seize them, and many operate from jurisdictions that don’t cooperate with U.S. law enforcement. The Manhattan DA can seize domain names and take down U.S.-hosted infrastructure, but sites can migrate to new domains or offshore hosts within hours.

The European dimension

The analysis of European politicians appearing on deepfake sites adds another layer to this. When the targets are elected officials rather than celebrities, the harm isn’t just personal. It’s a documented form of gendered political intimidation.

Nearly every politician identified in the research is a woman. That’s not a coincidence. It’s the pattern we’ve seen in every analysis of non-consensual deepfake pornography: the technology gets weaponized overwhelmingly against women, and it gets used to humiliate, silence, and discredit them.

Some European jurisdictions have stronger laws on the books than the U.S. does. The UK’s Online Safety Act, which went into full effect last year, creates criminal liability for hosting or sharing deepfake pornography. Germany amended its criminal code in 2024 to specifically cover AI-generated sexual imagery. France has prosecuted cases under existing image-based sexual abuse statutes.

But enforcement remains inconsistent. Cross-border cases are hard. Sites vanish and reappear under new names. And the technological barrier to creating convincing deepfakes keeps dropping.

What happens next

The Manhattan DA’s seizures won’t end deepfake abuse, but they do something important: they demonstrate that there are legal consequences for running these operations, at least if you’re within reach of a motivated prosecutor.

The question is whether other jurisdictions follow suit. The U.S. doesn’t have a comprehensive federal law criminalizing deepfake pornography. Some states do. New York’s laws on unlawful surveillance and promoting a sexual performance by a child have been used creatively in some cases, but they weren’t written with AI in mind.

Congress has considered several bills. None have passed. The closest we’ve come is the TAKE IT DOWN Act, which would create a federal right of action for victims and require platforms to remove deepfake pornography within 48 hours of notice. It’s been reintroduced multiple times. It keeps stalling.

Meanwhile, the European Union’s AI Act includes provisions requiring transparency for synthetic content and prohibiting certain uses of AI-generated imagery. Whether those provisions will actually curb deepfake abuse depends entirely on enforcement, and we won’t know that for another year or two.

For now, the Manhattan DA’s action is the most aggressive enforcement we’ve seen in the U.S. It matters. But 12 sites down, 148 more documented in the European analysis alone, and an unknown number operating elsewhere. The law is moving. It’s just not moving fast enough.