Seeing your own face in a video that was never filmed, or hearing your own voice say things you never said, brings a mix of disbelief, anger and helplessness. The first reaction is often to want everything gone immediately, at any cost. That is understandable, but acting on impulse can make things worse: evidence gets deleted, the spread is fuelled, or you respond in a way that makes the fake look more credible. What you need is a plan in a few clear steps.
Step 1: preserve everything before acting
Before flagging or requesting removal, gather the material. Once removed, the content may no longer be recoverable, and without it it will be hard to show what happened.
- Save the exact address of every page where the deepfake appears.
- Take screenshots showing the content, the profile that published it, the date and the address bar.
- If possible, save a copy of the video or image.
- Note the number of views, shares and significant comments.
- Write down in a document the date and the way you found out about it.
If the content is circulating on several platforms, do the same for each. Nothing complicated is needed: a tidy folder with files and notes is enough.
Step 2: flag it to the platforms
Every platform has rules on manipulated or deceptively generated content, as well as on impersonation, harassment and intimate images. Choose the most precise category:
| Type of deepfake | Category to use |
|---|---|
| Makes you say things you did not say | Manipulated media or false information |
| Poses as a profile of yours | Impersonation |
| Has sexual or intimate content | Non-consensual intimate images |
| Is used to mock or threaten you | Harassment or bullying |
In the text of the flag, state clearly that you are the person depicted, that the content was artificially generated or altered and that you never gave consent. If you have the original video or photo it was derived from, point to it: it is the strongest argument.
Step 3: do not amplify
It is natural to want to show everyone that the video is fake. But sharing it, even to debunk it, puts it in front of more people. Avoid:
- reposting the deepfake on your profiles;
- commenting under the original post, because comments increase its visibility;
- responding angrily to those spreading it.
Step 4: issue a sober denial
A short denial on your official channels helps the people who know you get their bearings. An effective wording:
"A video is circulating that shows me saying things I never said. It was created artificially. I have flagged it to the platforms. Please do not share it, and flag it if you see it."
No details about the content, no links, no direct accusations against people not identified with certainty.
If the deepfake concerns your work, also inform colleagues, clients or your employer before they hear about it from others: getting there first with a clear explanation greatly reduces the damage.
Step 5: consider legal protection
A deepfake can infringe several rights at once: reputation, image, personal data. Depending on the content, defamation, impersonation or, for intimate images, the unlawful dissemination of sexually explicit images may come into play. In Italy, the artificial intelligence law passed in 2025 also introduced a specific offence for spreading, without consent, AI-generated or altered content that causes unjust harm. At European level, the AI regulation sets transparency obligations for those who publish deepfakes.
With the material gathered in step 1, a lawyer can assess which route is most appropriate and whether to file a complaint. Remember that for some offences, such as defamation, there are strict deadlines for acting.
Step 6: keep the situation under watch
A deepfake removed from one platform may reappear elsewhere. In the following weeks, search your name periodically, including in combination with keywords related to the content, and set up an automatic alert for your name on search engines. Each new copy should be handled like the first: documented and flagged, citing the removals already obtained.
Looking after yourself
Being the victim of a deepfake has a real emotional impact, especially when the content is humiliating or sexual. Asking for help from people you trust or from a professional is not a weakness. Delegating the search for and flagging of copies to someone close can spare you from seeing the content over and over.
In short
- Preserve addresses, screenshots and copies before asking for any removal.
- Flag using the most precise category and state that you are the person depicted.
- Do not reshare the deepfake, not even to debunk it.
- Post a short denial and be the first to inform those who matter in your work.
- Consider legal protection with a lawyer and check over time that the content does not reappear.