More AI-generated child sexual abuse images found in the first six months of 2026 than all of 2025, IWF warns

Rise in AI child abuse imagery fuels call for urgent EU action

Expert analysts at the Internet Watch Foundation (IWF) have found more photorealistic AI-generated child sexual abuse images in the first half of this year than in the whole of 2025, including imagery depicting children under two years old.

The findings have prompted the IWF, Europe’s largest hotline working globally to find, assess and stop child sexual abuse images and videos online, to urge EU policymakers to agree on long-delayed legislation that would give tech companies lasting legal certainty to find and remove criminal child sexual abuse content on their platforms.

The new IWF data reveals a sharp rise in AI child sexual abuse imagery, with girls accounting for the vast majority of images identified and increasing numbers depicting younger children.

Between January 1 and June 30 this year, IWF analysts assessed 6,310  AI-generated images that met the legal definition of child sexual abuse. This is 40% more than recorded during the whole of 2025 when analysts identified 4,512 images.

Among the images where both age and gender were recorded (6,221), girls featured in 98% of AI-generated imagery identified during the first half of 2026. The number of unique images depicting girls rose substantially from 4,259 across the whole of 2025 to 6,094 by the end of June this year.

The data also shows an increase in AI sexual abuse imagery depicting younger children. During the first half of 2026, analysts identified:

• 2,534 images depicting children aged seven to 10;

• 2,369 images depicting children aged 11 to 13;

• 1,004 images depicting children aged three to six;

• 190 images depicting infants and toddlers under the age of two.

Together, imagery depicting children aged seven to 13 accounted for 79% of AI-generated images identified in the first half of 2026, up from 70% across the whole of 2025.

The IWF works to disrupt the spread of this content by converting confirmed criminal images into unique digital fingerprints known as ‘hashes’. These hashes can be shared with law enforcement and technology companies, enabling them to automatically detect and block known child sexual abuse imagery without storing or viewing the original images.

While hash-matching is a vital tool, it can only stop abuse content that has already been identified. The growth of AI-generated child sexual abuse material demonstrates why EU child protection legislation must enable the detection of both known and previously unseen content, says the IWF. New criminal content has no existing hash because it must first be identified before it can be given a hash and prevented from spreading.

The proposed Child Sexual Abuse Regulation represents ‘a real opportunity’ to close the gaps where services are not already detecting and removing child sexual abuse material. But a legal framework that focuses only on known material would leave the door open for offenders to rapidly create and spread high volumes of new AI material.

Online child sexual abuse has evolved significantly since the legislation was first proposed in 2022 and the EU needs a framework that can meaningfully respond to the scale of the challenge.

IWF Chief Executive Officer Kerry Smith said: “It is deeply concerning to see how quickly AI-generated child sexual abuse imagery continues to grow and spread online. If Europe is serious about protecting children, it needs a comprehensive Child Sexual Abuse Regulation that gives platforms the legal certainty to detect, prevent and respond to known and unknown child sexual abuse content.

“At a time when technology is making it easier than ever to create and distribute abusive images and videos at scale, Europe cannot afford to delay. The EU has already shown its resolve by strengthening criminal laws to tackle AI child sexual abuse through the recent Child Sexual Abuse Directive. The Regulation is a real opportunity to further strengthen online protections for children and allow tech companies to use proportionate tools to tackle the spread of the evolving threat. Without it, efforts to prevent abuse are weakened and children are less safe online.”

The IWF data also shows an increase in Category C child sexual abuse imagery, as graded under UK law, which can include sexualised posing and nude imagery. During the first half of 2026, analysts identified 5,557 Category C images which made up 88% of all AI images compared with 2,842 AI images (62%) in all of 2025. The IWF also found 350 Category A images and 403 Category B images.

Category A is the most severe category, and can include depictions of rape, bestiality or sexual torture. Category B can include non-penetrative sexual activity.