AI-Generated CSAM: What Parents and Platforms Should Know

Artificial intelligence can now generate photo-realistic images from a short description or an ordinary photo. The same technology is being abused to fabricate child sexual abuse material, sometimes called AI-generated or synthetic CSAM. This article explains what it is, how fast it is growing, why it harms real children even when an image is fabricated, and the practical steps parents and platforms can take. It does not name or describe any tool or method for creating this material.

What AI-generated CSAM is

AI-generated CSAM covers two broad categories. The first is imagery built entirely by a model with no specific real child in the frame. The second, and often more damaging in everyday cases, uses a real photograph of a real child and alters it to make the child appear nude or engaged in sexual activity. So-called "nudify" apps fall into this second group. They take an ordinary picture, a class photo, a holiday snapshot, a social media post, and fabricate an explicit version of it.

The Internet Watch Foundation, which finds and removes this material, warns that the most convincing AI-generated imagery can be visually indistinguishable from real photographs and videos, even for its trained analysts. That realism is part of what makes it so harmful and so hard to moderate.

How fast it is growing

The trend lines are steep. The IWF found 245 reports containing actionable AI-generated child sexual abuse imagery in 2024, up from 51 in 2023, a 380% increase. Those 2024 reports contained 7,644 images and videos, and 98% of the imagery depicted the sexual abuse of girls. The IWF also recorded a rise in the most severe category of abuse and a marked jump in video quality over the year.

The US picture is similar. The National Center for Missing and Exploited Children (NCMEC) received about 4,700 CyberTipline reports involving generative AI in 2023 and roughly 67,000 in 2024. NCMEC has said the offenders in many cases were people in the child's life with direct access to them, and that more than 275 direct victims of AI-generated CSAM were identified across 2024 and 2025.

Why it harms real children

A common excuse from people who create this material is that no real child was harmed because an image is synthetic. That claim does not hold up.

When a nudify app is used on a real child's photo, the child in the fabricated image is a real, identifiable person. NCMEC notes that this content can cause serious harm to victims, including harassment, bullying, and lasting psychological and emotional harm. The child has to live with explicit images of themselves circulating among peers, and often cannot fully remove them.

There is broader harm too. Fabricated imagery of known abuse victims can be produced on demand, forcing them to relive their abuse. And because synthetic material can look identical to photographs of real abuse, it strains the investigators and analysts who work to identify and rescue real children, making it harder to tell which images show a child who needs help right now.

Its use in sextortion and bullying

AI-generated imagery has become a tool for two fast-growing threats. In sextortion, an offender fabricates an explicit image of a young person and threatens to share it unless the victim pays money or sends real images. The threat works even when the starting image is fake, because the fear and shame are real. Financial sextortion has become a daily event. NCMEC received close to 100 financial sextortion reports a day through 2024.

The second threat is peer-on-peer bullying. Thorn's research found that 1 in 8 teens personally knows someone who has been targeted with an AI-generated deepfake image. In many school cases the person creating the fake imagery is a classmate. The result is humiliation and harassment that spreads quickly through a child's own social circle.

The legal status

In the United States, AI-generated CSAM is illegal at the federal level. The Department of Justice has been clear that there is no distinction under the law between abuse material involving a real victim and generative AI depictions for the purpose of prosecution. Imagery that alters a real child's photo is charged under child sexual abuse material laws. Wholly synthetic imagery has been prosecuted under federal obscenity laws, which historically carry different penalties. Lawmakers have moved to close that gap. The ENFORCE Act, introduced in 2025 and passed by the Senate in December 2025, is intended to standardize how these offenses are charged and penalized.

What parents can do

• Talk with your child before it happens. Let them know that fake explicit images exist, that being targeted is never their fault, and that they can come to you without getting in trouble.
• Reduce the raw material. Review who can see and download your child's photos, and tighten privacy settings on their accounts and yours.
• If your child is targeted, do not have them pay or send anything. Save evidence such as usernames and messages, and report it.
• Report to the NCMEC CyberTipline at report.cybertip.org or 1-800-843-5678. Reports are reviewed by people, not processed automatically.

What platforms can do

• Treat AI-generated CSAM as CSAM in policy and enforcement, including altered images of real children.
• Restrict features that let users fabricate explicit images of identifiable people, and act on the apps and accounts that offer them.
• Make reporting fast and give clear options to flag suspected AI-generated content, then route reports to trained human reviewers.
• Support victim removal, including hash-matching and cooperation with NCMEC and the IWF so known imagery can be found and taken down.

The technology is new, but the duty is familiar. A fabricated image of a real child is a real harm to that child, and it is treated as such under US law.