How to Report AI-Generated Explicit Images of a Minor

Where do you report AI-generated explicit images of a minor?

In the United States, report it to the National Center for Missing and Exploited Children (NCMEC) through the CyberTipline at report.cybertip.org, or by phone on 1-800-843-5678. NCMEC reviews each report and makes it available to the law-enforcement agency with jurisdiction. The FBI's Internet Crime Complaint Center (ic3.gov) accepts reports too, and NCMEC's own guidance lists both routes.

Call local emergency services first if a child is in immediate danger, is being threatened, or has talked about self-harm. Schools, clinics, youth organizations and online platforms usually carry separate mandatory-reporting or safeguarding duties, so the designated safeguarding lead should be told the same day.

A CyberTipline report is stronger when it carries the details investigators can act on: the URL where the image appeared, the platform, usernames and profile links of whoever posted or sent it, dates and times, and any demands or threats. Keep the report confirmation number with your own notes.

• CyberTipline (NCMEC): report.cybertip.org or 1-800-843-5678, available 24 hours.
• FBI Internet Crime Complaint Center: ic3.gov, especially where money or more images are being demanded.
• The platform's own reporting tool, used in addition to the CyberTipline, never instead of it.
• Outside the US, the national hotline or police unit that handles child sexual abuse material.

Why you should never download, screenshot or forward the image

Every new copy of a sexual image of a child is a new piece of illegal material, even if it was made by an AI tool and even if you are trying to help. The FBI's public service announcement of 29 March 2024 states that federal law prohibits producing, distributing, receiving and possessing child sexual abuse material, and that this includes realistic computer-generated images. Parents, teachers and moderators who save the file as evidence can create legal exposure for themselves.

Record the context instead. Write down or capture the page address, account names, message timestamps and the text of any threats, without capturing the image itself. If the material arrived on a child's phone, leave it where it is; investigators can collect it properly. NCMEC's Take It Down service says the same thing in plain terms: do not send, share or download an image to submit it.

How often are AI-generated child sexual images reported?

Reports have grown roughly a hundredfold in three years. NCMEC began tracking generative-AI-related CyberTipline reports in January 2023 and counted 4,700 that year, 67,000 in 2024 and more than 400,000 in 2025. The largest 2025 category, over 145,000 reports, involved users altering an existing abuse file with AI. Another 30,000 or more involved uploading a file together with text prompts to generate new material.

NCMEC staff labelled more than 158,000 images and videos as AI-generated between January 2023 and December 2025, while electronic service providers flagged just over 11,000 over the same period. That gap matters: platforms often do not know whether a file is synthetic when they report it, and NCMEC identified more than 275 direct victims of AI-generated abuse material in 2024 and 2025. Behind many of these files is a real child whose ordinary photo was altered.

Getting the image removed with Take It Down

Take It Down is a free NCMEC service for nude, partially nude or sexually explicit images or videos taken when the person shown was under 18. The tool computes a hash, a numeric fingerprint, on the device that already holds the file. Only that hash goes to NCMEC, and participating platforms use it to find and remove matching copies. The image itself stays on the device and is never uploaded.

The service has limits worth knowing before you rely on it. It works only with platforms that have joined, and only on public or unencrypted services. NCMEC notes that platforms may have limited ability to remove content posted in the past, and an AI-altered image that has been re-edited or cropped may no longer match its hash. Adults whose images date from age 18 or later are directed to StopNCII instead.

In the US, the TAKE IT DOWN Act, signed on 19 May 2025, adds a separate legal lever. Covered platforms had until 19 May 2026 to set up a notice-and-removal process and must remove qualifying intimate images, including digital forgeries, within 48 hours of a valid request.

Supporting the child while the report is handled

Tell the child plainly that they are not in trouble. Many victims of AI-altered images never posed for anything; a classmate or stranger took an everyday photo and ran it through an app. The FBI alert specifically mentions teenagers creating nude images of classmates from clothed photos. If someone is demanding money or more images, do not pay, do not reply, and do not delete the conversation.

Limit how many times the child has to look at or describe the material. One trusted adult should coordinate with the school, the platform and investigators so the child is not interviewed repeatedly. NCMEC offers crisis counselling referrals and Team HOPE, which connects families with others who have been through the same thing.

• Secure the child's accounts and review who can contact them, after the evidence is recorded.
• Watch for renewed contact from new accounts, a common pattern in sextortion.
• Agree with the child what school staff need to know, and avoid any public statement that names them.

Where deepfake detection helps, and where it has no place

Families and schools do not need a detector to report. The reporting duty is the same whether an image is real, altered or fully generated, and running a file through any online tool creates exactly the kind of copy the guidance warns against. Detection belongs with organizations that already hold media under legal authority: trust and safety teams at platforms, law-enforcement units and the analysts who support them.

For those teams, the useful question is often which parts of a file were altered and how confident the analysis is. NeuroDetect, Neural Defend's forensic media workspace, keeps each file inside a case, records the model and version used, and shows explainability heatmaps down to regions and frames. Its downloadable forensic report lists observations, uncertainty, provenance and analyst notes, and the analyst makes the final call. That record can support triage and victim identification work, for example showing that a face was pasted onto another body.

Neural Defend supports on-premises and air-gapped deployment for teams that cannot send sensitive material to a cloud service, and holds ISO/IEC 27001, ISO/IEC 27701 and SOC 2 Type II certifications. Any deployment involving child sexual abuse material has to run under the organization's legal authority and agreed handling rules, scoped with Neural Defend during evaluation.