What to Do If Someone Made a Deepfake of You

If you've found a fake or manipulated image or video of yourself online — or you're afraid one is about to be shared — you are not powerless, and you have not done anything wrong. This page walks through what to do, in order, using free services that are built for exactly this situation.

Everything here works whether the image is AI-generated or a real photo shared without your consent. The removal tools treat both the same way.

If you are in immediate danger, being threatened, or a minor is involved, treat this as urgent. Skip to If you're being threatened or blackmailed and If the person in the image is under 18.

First: a few things that are true

• This is not your fault. Being targeted with a deepfake or a leaked image is something done to you.
• You don't have to look at it repeatedly or handle this alone. The steps below are designed so you do as little as possible.
• You do not need to send the image to anyone to get it removed. The main tools create a fingerprint of the image on your own device — the image itself never leaves your phone or computer.

Step 1: Don't delete everything — but don't spread it

Two instincts to resist in the first few minutes.

Don't delete every trace in a panic. You may want proof later — for a report to the platform, a school, an employer, or the police. Keeping a record is covered in Step 4.

And don't download, re-share, or send the image to friends to ask "is this me?" — even to warn people. Forwarding it, including of a minor, can carry its own legal risk and spreads it further. You can get it removed without circulating it.

Step 2: Get it removed — free, without uploading the image

The most effective free removal tools work by hashing: your device turns the image into a unique digital fingerprint, and only that fingerprint is sent onward. Participating platforms use it to find and block matching content. The image never leaves your device, and no one at these services views it.

Which service you use depends on how old the person in the image is — or was when it was taken.

If you are 18 or older: StopNCII.org

StopNCII.org is a free service run by a UK-based non-profit (the Revenge Porn Helpline / SWGfL). It's for adults dealing with non-consensual intimate images, including AI-generated and deepfake content.

How it works, briefly:

• You select the image or video on your own device.
• Your device generates a hash. The hash is sent — the image is not, and it stays with you.
• You get a case number and PIN. Save both somewhere safe; they can't be recovered if lost, and you'll need them to check or withdraw your case.
• Participating platforms look for matches and remove content that breaks their intimate-image-abuse rules. The service keeps checking over time.

What it can't do: it can only act on participating platforms, not the entire internet. It's still the best single starting point for adults.

If the person in the image is under 18: Take It Down

If the image is of someone who is currently under 18, or was under 18 when it was taken, use Take It Down, run by the U.S. National Center for Missing & Exploited Children (NCMEC). It's free and anonymous, works the same on-device hashing way, and explicitly covers AI-generated images. A parent or trusted adult can help a young person do it.

Quick rule: 18 or over, use StopNCII. Under 18, use Take It Down. If in doubt for a minor, start with Take It Down.

Step 3: Report it to the platform directly

Alongside the hashing tools, report the content to the specific site or app hosting it. Most major platforms now have a dedicated category for non-consensual intimate images and synthetic or manipulated media, and reporting it yourself is often the fastest route on that one platform. Look in the report menu for options mentioning "non-consensual," "intimate image," "deepfake," or "synthetic media."

If the content is on a smaller site with no real reporting process, the hashing tools in Step 2 and the support services in Step 5 are the better path.

Step 4: Keep a record, in case you need it

Before content comes down, it's worth preserving proof — quietly, without sharing it. Note the URLs where it appears, the dates you saw it, and any usernames or messages involved. A record helps if you later go to the platform, a school or workplace, or law enforcement.

This is also where confirming whether an image is AI-generated can help. A verification result — showing that an image carries the fingerprints of a generative model, or that its provenance data is missing or inconsistent — can be a useful part of your record when you're explaining the situation to someone who needs to act. You can check an image at witness.vision/scan; it's free and the result is yours to keep.

A verification result is a helpful first signal, not certified forensic evidence. If a formal legal process is involved, a qualified digital forensics expert can produce a court-suitable analysis.

Step 5: You don't have to do this alone

Removing an image is practical. The rest of it — the shock, the anxiety, the feeling of exposure — is not something to white-knuckle by yourself. These services exist specifically for image-based abuse and are staffed by people who deal with it every day:

• Cyber Civil Rights Initiative (CCRI) — a helpline and resources for image-based abuse. cybercivilrights.org
• StopNCII's support resources — stopncii.org/resources-and-support
• NCMEC (for anyone under 18 or their families) — missingkids.org

If you're a student and this involves your school, a trusted teacher or counselor can help you navigate it — you don't have to explain everything at once.

If you're being threatened or blackmailed

If someone is demanding money or images and threatening to share a deepfake ("sextortion"):

• Don't pay, and don't send more images. Paying or complying almost always leads to more demands, not fewer.
• Stop engaging with the person, but don't delete the conversation — it's evidence.
• Report it. In the U.S., the FBI takes sextortion reports at ic3.gov and 1-800-CALL-FBI. If a minor is involved, contact NCMEC's CyberTipline at report.cybertip.org.
• If you feel you're in immediate physical danger, contact local emergency services.

If the person in the image is under 18

This is treated with the highest urgency, and there are dedicated, free services:

• Take It Down (takeitdown.ncmec.org) to hash and block the image across participating platforms.
• NCMEC CyberTipline (report.cybertip.org) to report it.
• A parent, guardian, or trusted adult can help with any of this. It's free and anonymous.

Do not download or forward the image to report it — select it in the tool directly, which hashes it on the device without it leaving.

Where Witness fits — and where it doesn't

We build image and video verification. Witness can tell you, for free, whether an image or video shows signs of AI generation, and it reads provenance data (Content Credentials) when present — useful for confirming what you're dealing with and for your own records.

Witness does not remove content from the internet. For removal, the free services above — StopNCII for adults, Take It Down for minors — are purpose-built, run by dedicated organizations, and more effective at it than any detection tool could be.

If confirming whether something is real would help, you can scan an image or video here. Otherwise, start with Step 2 — that's the part that gets the content down.

This guide is informational and not legal advice. Laws on non-consensual and synthetic intimate imagery vary by country and region, and services and their participating platforms change over time — check each linked service for its current details.