There is a deepfake of you online. Here is what to do.

How to tell if you are targeted

A deepfake is content generated or altered by AI that shows you in a situation that never happened. The most common signs:

• a friend, colleague or follower sends you a “strange” link or screenshot;
• unusual messages arrive, sometimes with a demand for money;
• searching your name or photo brings up pages you do not know;
• an unknown account posts images of you that you never took.

If you never filmed that scene or said those words, it is fake. You do not have to “prove” it technically to report it: the lack of consent is enough.

The first 7 steps

1. Get safe. If you are threatened, call your local emergency number. If you are being blackmailed, do not pay and do not negotiate: paying almost never gets the content removed.
2. Keep evidence before the content moves:
   1. Copy the exact address (URL) of every page, post or account involved.
   2. Take a screenshot showing the content, the address and, if possible, the date. Do not download or share the content itself.
   3. Note who posted it: account name, handle, date posted, any useful comments.
   4. Keep any messages if someone contacted you (threats, demands for money). Do not delete them.
3. Report the content and the account on the platform that hosts it, with the most precise reason: non-consensual intimate imagery, manipulated or AI-generated media, impersonation or harassment.
4. Get it out of Google. Google has a dedicated form for fake sexual content, and others for personal information. Read our Google guide.
5. Escalate if the platform refuses or stays silent: use the internal appeal (mandatory for large platforms in the EU), and in the US, cite the TAKE IT DOWN Act for intimate images, which gives platforms 48 hours to act.
6. Decide whether to report it to the police. It is your choice. Victim support services and helplines can help you for free, and a lawyer can advise you.
7. Watch for copies. Search your name, run reverse image searches: every new copy is handled like the first one.

Why reports often fail

Reports rarely fail because the request is unfounded. They fail because:

• the wrong reason is picked (“inappropriate content” instead of “impersonation” or “non-consensual intimate imagery”), so the request lands in an automated queue;
• the platform asks for proof of identity and the request stops there;
• the content has been copied elsewhere: it is removed in one place and stays up in ten others;
• nobody follows up or uses the appeal that large platforms must offer under the EU Digital Services Act.

What happens if you leave it

• The content is copied to other sites, then indexed by search engines under your name.
• It resurfaces at a job interview, a first date or a contract negotiation.
• It can fuel harassment or the exposure of your personal details.
• The mental toll grows: talk about it, and do not hesitate to ask for professional support.

What the law says

More and more countries treat non-consensual deepfakes as a crime. In the United States, the TAKE IT DOWN Act makes publishing non-consensual intimate images, including AI-generated ones, a federal crime. In the United Kingdom, sharing an intimate image without consent is an offence, deepfakes included. In the European Union, platforms must handle reports of illegal content under the Digital Services Act, and in France sharing a deepfake of someone without consent is punishable by up to two years in prison when done online. Read our guide to deepfake laws.

Reporting to the police is your right, and it is the only way to have the person behind it identified and prosecuted. It is not required to get content taken down. A lawyer in your country can tell you whether it makes sense in your case, and our evidence file can be attached to your report.