AI-Generated Imagery

Deepfakes: everything you need to know

AI-generated images and videos depicting real people without their consent are now one of the fastest-growing categories of online harm. The 2026 TAKE IT DOWN Act covers them explicitly. Here is how to understand, identify, and remove them.

90% of deepfake content is sexually explicit
48 hr statutory removal window under federal law
2026 TAKE IT DOWN Act explicitly covers AI content

Understand deepfakes

What deepfakes are and how the technology works.

Explainer · 8 min
What Is a Deepfake?
A plain-English explanation of what deepfakes are, how they are created, who gets targeted, and what the law now says about them.
Technology · 10 min
How Are Deepfakes Made?
The AI technology behind face-swaps, what data is needed, and why biometric scanning finds what reverse image search misses.
Research · 7 min
Deepfake Statistics 2026
Volume, growth rates, most targeted demographics, and the platforms where deepfakes spread.
Technology · 9 min
How to Tell If an Image Is a Deepfake
The visual tells in AI-generated images and video, detection tools that work, and what to do when you find one of yourself.

Take action

Legal rights, removal steps, and what to do right now.

Legal Guide · 9 min
Is Creating a Deepfake Illegal?
Federal law, 26 state laws, UK and EU and Australia: the complete legal picture in 2026.
Legal Guide · 8 min
Does the TAKE IT DOWN Act Cover Deepfakes?
Yes. The 2026 version explicitly covers AI-generated content. Which platforms are bound and how the 48-hour window applies.
Crisis Guide · 11 min
How to Remove a Deepfake From the Internet
Step-by-step: find every location, file removal notices under the TAKE IT DOWN Act, and get every match delisted across 200 or more platforms.
Crisis Guide · 9 min
Deepfake Porn: How to Find It and Force Its Removal
AI-generated intimate imagery is now a federal offence. How to detect it, why standard search fails, and how to compel removal.

What the law says

The 2026 TAKE IT DOWN Act amended 47 U.S.C. to add Section 223a, which creates a direct federal obligation for covered platforms to remove non-consensual intimate imagery within 48 hours of receiving a compliant notice. The Act's definition of protected content explicitly includes synthetically generated or altered visual depictions, meaning AI-generated content is covered in exactly the same way as authentic photographs.

Prior to 2026, deepfakes occupied a grey area in US federal law. Some states had specific deepfake statutes, but there was no unified federal removal mechanism. The TAKE IT DOWN Act resolved this by creating a single federal pathway applicable to all covered platforms, regardless of where they are hosted.

Why biometric scanning is required

Deepfakes are generated from a reference image, typically scraped from public social media. The generation process changes the pixel-level data of the image entirely, which means standard reverse image search, which matches pixel patterns, returns no results. What remains constant is your facial geometry. Biometric scanning maps the geometry of your face and searches for matches regardless of whether the found image is an authentic photograph or AI-generated content.

Frequently asked questions

What is a deepfake?
A deepfake is AI-generated or AI-altered content depicting a real person in a context they were never in. Deepfakes are created using generative adversarial networks or diffusion models trained on public images. Over 90 percent of deepfake content online is sexually explicit. The 2026 TAKE IT DOWN Act covers AI-generated intimate imagery in exactly the same way as authentic photographs.
Does the TAKE IT DOWN Act cover deepfakes?
Yes. The 2026 version explicitly covers synthetically generated or altered visual depictions. A deepfake of your face in a sexually explicit context is treated identically to an authentic photograph under federal law. Covered platforms have 48 hours to comply with a valid removal notice from the moment of receipt.
Can standard reverse image search find deepfakes of me?
No. Standard reverse image search matches pixel data and cannot identify AI-generated content that uses your facial geometry without copying the original pixels. Biometric face scanning identifies matches based on facial geometry regardless of whether the image is authentic or AI-generated.
How long does deepfake removal take?
Covered platforms have a 48-hour statutory window from receipt of a compliant Section 223a notice. Major platforms including Meta, X, Reddit, and TikTok typically respond within this window. Smaller platforms and anonymous hosting services may take longer.

Found a deepfake of yourself?

Upload a photo. We biometrically scan 200 or more platforms and dispatch 48-hour removal notices to every match.

Start removal scan
Swiss Privacy  ·  Erase Images Erase Your Images Online for $89  ·  Verified 47 U.S.C. § 223a
Upload your image