If an AI Deepfake Targets a Child, Do These Things First
Matthew Leo · Published October 9, 2026 · Guides
The number of AI-generated child sexual abuse images identified by the Internet Watch Foundation is rising sharply. The organization says it found 6,310 images that met the legal threshold in the first half of 2026, compared with 4,512 in all of 2025.
Those figures describe material assessed by the IWF, not the total amount circulating online. They still show why families and schools need a response plan before an incident lands in a group chat.
If a fake sexual image targets a child or teenager in Canada, the priority is the child's safety. It is not proving that the image is fake to every person who received it.
Do not download or forward the image
Do not save extra copies, resend the file for advice or ask the young person to show it repeatedly. Moving sexual material involving a minor can spread the harm and create legal risk.
Preserve information around the post instead. Record the account name, page address, date and time, platform, conversation name and any non-explicit messages that show threats or distribution. If taking a screenshot would capture the sexual image itself, use the reporting service's instructions before doing so.
Tell the child what happens next
A young person may fear punishment, school gossip or losing access to their phone. Make clear that the priority is stopping the spread, not blaming them. Avoid promising that every copy can be erased; promise instead that an adult will help with reporting, support and follow-up.
If there is an immediate threat to the child's safety, call 911. Kids Help Phone is available across Canada at 1-800-668-6868, and young people can text CONNECT to 686868.
Report it through Canadian channels
- Report the account or post to the platform. Use the category for child sexual exploitation, non-consensual intimate imagery or impersonation, whichever best matches the service's current menu.
- Submit a report to Cybertip.ca. Canada's national tipline can receive reports concerning online sexual exploitation of children.
- Use NeedHelpNow.ca for removal steps. Its guidance covers platform reporting and takedown requests. Platform menus change, so use the current instructions on the site rather than an old screenshot.
- Contact local police when a child is threatened, extorted or repeatedly targeted. Tell them whether the image appears synthetic and where it was shared, without circulating another copy.
A school should not handle the incident as ordinary teasing. The Canadian Centre for Child Protection has a school response guide for non-consensual distribution of intimate images that covers both authentic and synthetic material.
Why speed matters, and what it cannot solve
The IWF says 79 per cent of the AI-generated images in its first-half dataset depicted children aged seven to 13. It also notes a technical limitation: hash matching can help platforms detect known images only after analysts have identified and fingerprinted them. Newly generated material may not match anything in an existing database.
That is why a report still matters even when a platform claims to scan uploads. Human reports can identify a new image, connect related accounts and start a removal process.
What parents and schools should avoid
- Do not conduct a public investigation in a class or parent group.
- Do not name an alleged creator without verified evidence.
- Do not require the targeted child to confront other students.
- Do not treat a synthetic image as harmless because no camera was used.
- Do not rely on a single removal request if the image appears on several services.
Cybertip.ca's deepfake guidance stresses that fake images can cause real harm. The useful distinction is not between “real” and “fake” abuse. It is between verified facts, unverified claims and the concrete steps that reduce distribution and support the child.
Tags: AI deepfakes, Cybertip.ca, NeedHelpNow.ca, image-based abuse, child protection