Someone made fake nude images of you with AI
Someone took an ordinary photo of you — a holiday snapshot, a work headshot, a picture off your own profile — ran it through an app, and produced an image that looks like you naked. Maybe a friend sent you a link. Maybe it turned up in a group chat, or on a site you had never heard of. It never happened, and you never consented to any of it, and yet there it is with your face on it.
People in this situation often describe the same first reaction: a frozen kind of embarrassment, as though they owe someone an explanation before they are even allowed to ask for help. You don't. Nothing you did created this, and the apps that produce it are now so cheap and so widely available that being targeted tells the world nothing about you at all. What follows is the sequence I would walk you through, in the order that actually gets results.
The image is fake. What it can cost you is not
It helps to name what this is. Synthetic intimate images — sometimes called deepfake nudes, or the output of “nudify” apps — are made by feeding a clothed photo into software that generates a fabricated nude body under a real face. No private photo of you ever needed to exist. That is why the usual advice about being careful what you send simply doesn't apply here, and why anyone can be targeted, including people who have never taken an intimate photo in their life.
It also means you are dealing with something the law and the major platforms now treat as image abuse, not as a prank or a doctored joke. That distinction matters, because it is what gives you leverage in the steps below.
Capture it first, even though you want it gone
There is a hard truth in the order of operations here: the successful takedown erases the proof along with the image. Once a platform removes the content, the record that it ever existed goes with it, and that record is exactly what you would need to identify who posted it, or to bring this to police or a lawyer. So before you report anything, capture it. Screenshot the post or page so that the full web address, the posting account or username, the date and time, and any caption or comments are all visible in the frame. If it is moving through a group chat, screenshot the chat with the sender's name and timestamps showing. Note everywhere else you have seen it. Looking at it long enough to do this is genuinely unpleasant, and it is also the step that keeps your options open. Our guide on how to preserve evidence in the first 24 hours covers the capture details.
Don't message the poster, don't argue in the comments, and don't ask friends to pile onto the account. Contact tends to escalate things and can push someone into reposting elsewhere before a takedown lands.
You now have a legal right to have it removed
Most people do not realize the law has caught up to this, so here it is. The TAKE IT DOWN Act covers AI-generated intimate images explicitly, not only real ones, and since 19 May 2026 the FTC has been enforcing it. Covered platforms — broadly, sites and apps that host user content — must give you a way to request removal, and must take the image down, along with known identical copies, within 48 hours of a valid request.
So go to the platform's reporting flow and use it, in your own name as the person depicted. Be plain in the request: the image depicts you, it is a non-consensual intimate image, it is a digital forgery, and you are requesting removal. Keep a copy of the ticket or reference number and the date you filed it. If 48 hours pass and nothing happens, or the site offers no removal process at all, you can file a complaint with the FTC at TakeItDown.ftc.gov, which exists specifically for platforms that ignore valid requests. Also submit the web addresses to Google's removal request form for explicit imagery, which takes the pages out of search results even when the host itself won't cooperate.
Get ahead of the copies
StopNCII.org is the single most useful tool here, and it accepts synthetic and deepfake images as well as real ones. It works by creating a digital fingerprint, or hash, of the image on your own device — the picture itself never gets uploaded anywhere. That hash is shared with participating platforms, which then block matching uploads before they go live. It is free, it takes a few minutes, and it is the only step that works ahead of the spread rather than behind it.
If the person depicted is under 18, use NCMEC's Take It Down service instead, which does the same hash-matching for minors, and treat the situation as child sexual abuse material regardless of who made it or how it was made. Our post on financial sextortion of teens covers what parents can do when a young person is the one targeted.
If someone is demanding money
Sometimes fabricated images arrive with a threat attached: pay, or this goes to your employer, your family, your followers. That is sextortion, and the answer does not change because the images are fake. Paying marks you as someone who pays and tends to bring a second demand rather than an ending. Keep the messages, stop replying, and work the takedown and reporting steps instead. We go through this in more detail in what to do when someone threatens to share your private photos.
Reporting it as a crime
Publishing non-consensual intimate images, including digital forgeries, is a federal crime under the same Act, and most US states have their own laws covering synthetic sexual imagery. File a report with IC3.gov and with your local police department, and bring your saved screenshots with you rather than describing the content from memory. Ask for a report number even if the officer seems unsure what to do with it — that number is what you will need for a school, an employer, a protective order, or a civil claim later.
If the account behind it is anonymous and you want to know who is actually running it, that is a separate question from getting the content removed, and the two can run in parallel. Do the takedowns first; identification depends on records held by platforms and payment providers, and those records survive long after a post is deleted.
Let me be honest about that identification, because these cases split two ways. In a lot of AI-nude situations the person who made the image is not a stranger at all — it is a classmate, an ex, someone in the same friend group, and their choice of who to target is itself a short suspect list. Those cases are frequently solvable, because a person you already half-suspect leaves the kind of trail you can follow. The harder version is a genuinely anonymous poster who dropped the image on an offshore forum built to shield exactly this, with no reused account and no payment attached. There the open-source path can stall, and the realistic route is a preserved-evidence package handed to police who can compel the records. Knowing which of the two you are in changes everything about what is worth spending.
One more thing worth saying plainly: you do not owe anyone an explanation for an image you did not make, did not pose for, and did not consent to. The people who need to know can be told in one sentence. Everything else is a technical problem with a well-worn set of steps, and it is one you can work through.
— Lisa, GCCI