Someone Made a Deepfake of Me, What Should I Do?

Someone sent you a link, or tagged you, or just said "have you seen this?" And now you're looking at a video or image of yourself doing or saying something you never did.

The first minutes after finding a deepfake of yourself are disorienting, and the instinct to immediately post a denial, message the person who shared it, or start deleting things is completely understandable. It's also usually not the most useful first move. Here's what actually helps, in order.

First: don't engage publicly yet

It's tempting to respond immediately, to comment, to post a denial, to confront whoever shared it. Give that instinct a few minutes before acting on it. A public reaction posted before you understand what you're dealing with can lock in a version of events before you've had a chance to establish the facts, and it can also alert whoever created the content before you've preserved evidence of it.

This doesn't mean staying silent forever, it means sequencing your response so the facts come first.

Preserve what you're looking at before it can disappear

Get a second, calmer look before assuming the worst

Not everything that looks unusual is a deepfake, and not every deepfake is as sophisticated as it first appears. Before escalating, it can help to have someone you trust take a second look, sometimes context (a joke account, an obvious parody, a low-effort edit) becomes clearer once the initial shock passes. This isn't about minimizing something genuinely harmful; it's about making sure your next steps are aimed at the actual situation, not the version of it that felt most alarming in the first thirty seconds.

Secure your own accounts

While you're working through the above, take a few minutes to check your own accounts , email, social media, anything connected to your identity. Deepfakes sometimes accompany account compromise or are created using material pulled from an account that's already been accessed without permission. Updating passwords and enabling multi-factor authentication where you haven't already is a small step that closes off one way the situation could get worse.

Report it through the platform's actual reporting channel

Most major platforms have specific policies against synthetic or manipulated media impersonating a real person, and a specific reporting category for it, which usually gets handled faster and more seriously than a generic "inappropriate content" report. Use the category that actually matches what happened (impersonation, synthetic media, non-consensual imagery, harassment) rather than the first option in the menu.

If the content involves intimate or sexually explicit imagery, real or AI-generated , there are purpose-built tools for this specifically: StopNCII.org for adults, and the National Center for Missing & Exploited Children's Take It Down tool if you were under 18 when the image was created. Both work by generating a digital fingerprint (a "hash") of the image on your own device, the image itself is never uploaded or seen by anyone, which participating platforms then use to detect and remove matching content. Eterna's Verification Methodology covers documentation in more depth if you need it for a platform report or beyond.

Know that you're not required to prove a negative

When to bring in outside help

Most single-platform incidents can be handled by following the steps above. Outside help becomes more useful when the content is spreading across multiple platforms at once, when it's being used as part of a scam or extortion attempt, when you're a public figure or executive and the content is gaining traction quickly, or when you need evidence preserved and documented to a standard that would hold up if legal action becomes necessary. That's the point where professional monitoring and response, the kind that verifies findings against a clear standard before treating anything as confirmed, and escalates only with your authorization, tends to matter more than doing it alone.

What happens next

Once you've preserved evidence and reported it, the Impersonation Response Guide walks through the fuller sequence, documenting the case, deciding what to communicate and to whom, and monitoring for the content reappearing elsewhere, which does happen and isn't a sign that anything went wrong the first time.