400,000+ Reports of AI-Linked Child Exploitation in 2025 — and the Raw Material Can Be the Family Photos Already Online
Back to blog
Family Safety·NOVEMBER 12, 2026

400,000+ Reports of AI-Linked Child Exploitation in 2025 — and the Raw Material Can Be the Family Photos Already Online

The Blaick Team 5 min read

The National Center for Missing and Exploited Children (NCMEC) received more than 400,000 reports involving generative AI and child exploitation in 2025. Investigators told ABC13 this month that AI tools can build sexualised images of a real child from photos already online, with no access to private pictures and no contact with the child.

The number to hold: 400,000+. Not every report was confirmed as AI-generated abuse after investigation, but the direction is clear, and the starting point is often an ordinary photo that a proud relative or friend posted.


NCMEC reports, 2025
400,000+
Reports involving generative AI and child exploitation.
What it needs
Only public photos
No private images and no contact with the child.
Skill required
Very little
A Virginia Tech researcher says it needs little technical expertise now.
Enforcement gap
Hard to prosecute
Often the person who made or shared the images can't be identified.

Sources: ABC13/WSET, "AI-generated child sexual abuse images pose growing threat to children" (November 2, 2026), citing NCMEC, Virginia Tech and the Southern Virginia Internet Crimes Against Children Task Force.


The Photo Is the Starting Point

For years, online safety advice focused on strangers contacting children. This is different: the image can be made without your child ever being contacted. Victim Maggie Southall-Bartz put it this way: "It's not just strangers anymore; they're finding your kids through online forums."

What investigators say

Clear, high-resolution photos make the best source material. Investigators advise limiting those, avoiding swimwear pictures, and keeping accounts private, while accepting that private settings do not remove the risk.

"You can do this without having much technical expertise at this point." — Bimal Viswanath, Virginia Tech

The Law Is Moving, the Companies Are Quiet

Legal picture

In Virginia, creating or possessing AI-generated abuse material using a real child's likeness is a felony. But prosecution depends on identifying the person responsible, and that is often the hardest step.

Company response

Meta, Google and xAI told ABC13 they have safeguards but declined to share details. Other companies contacted did not respond or agree to an interview.


What You Can Do This Week

Go private, then prune Make accounts private and remove followers you don't actually know.
Rethink what you post Skip sharp, high-resolution face shots and swimwear photos, and ask relatives to do the same.
Keep talking As one investigator said, check in "not because you don't trust your kids." Make sure they know they can tell you without fear of punishment.
Report it, even if it's fake Contact local police or the NCMEC CyberTipline about suspected AI-generated abuse images.

The photos are already out there; what you control is what goes up next and who can see it. A quick review of your accounts is one of the most effective things a parent can do right now.


Protect what your family shares online

Start a free 14-day Blaick trial — no credit card required.

Blaick helps you understand how AI touches your child's digital life, so you can act early.

Start free trial →
Cancel anytime · Trusted by security-conscious families · Built on enterprise AI security