Through generative AI, creating images or even videos will only take seconds thanks to simple, no-fluff prompts.

While the technology has legitimate uses in art, education, entertainment, and other fields, it is also being exploited to produce AI-generated child sexual abuse material (CSAM).

New figures from the Internet Watch Foundation (IWF) show how quickly the problem is developing. Between January and June 2026, IWF analysts assessed 6,310 AI-generated images that met the legal definition of child sexual abuse.

The figure was already higher than the total recorded during all of 2025 in the specific category covered by the latest IWF assessment. The material was overwhelmingly associated with girls, with many images depicting children between 7 and 13 years old.

How Quickly Is AI-Generated Child Abuse Material Increasing?

The latest IWF figures point to a significant acceleration.

The organization assessed 6,310 AI-generated images during the first six months of 2026. The Guardian reported that this represented roughly a 40% increase compared with the slightly more than 4,500 images identified during the comparable measurement of 2025.

The broader IWF data also show that AI-generated abuse material has expanded rapidly across different formats. In 2025, the organization identified 8,029 AI-generated images and videos depicting realistic child sexual abuse, a 14% increase from the previous year. It also recorded 3,443 AI-generated abuse videos, compared with only 13 in 2024.

The rise is significant because generative AI can reduce the technical knowledge needed to manipulate images. Instead of relying on traditional editing skills, offenders can increasingly use automated tools to produce or alter material.

Children Are Also Reporting AI-Manipulated Images of Themselves

The problem extends beyond completely synthetic images.

Children are increasingly reporting that photographs of themselves have been digitally altered to create fake sexual imagery.

The IWF’s Report Remove service recorded 420 reports involving imagery that children believed had been faked during the first six months of 2026. That was already higher than the 397 such reports recorded throughout 2025.

The figures include content suspected of being manipulated through AI, so-called nudification tools, computer-generated imagery, or other forms of image editing. They do not mean that every report involved generative AI.

This distinction matters because digitally manipulated abuse can begin with an ordinary photograph. A picture posted publicly can potentially be copied and altered without the child’s knowledge or consent.

For victims, the fact that an image is fake does not necessarily make the experience harmless. The image can still be circulated, used for intimidation, or associated with the child’s identity.

Why AI-Generated CSAM Creates a Different Challenge

Traditional child sexual abuse material involves the exploitation of real children during the production of an image or video. AI-generated material can introduce a more difficult hurdle never encountered before.

Some AI-generated images are entirely synthetic, while others can involve the likeness of real children. The IWF has warned that offenders can use real photographs as inputs or use existing abuse material to develop models capable of generating further content.

That means AI can potentially extend the harm beyond the original image.

A child who has already been photographed or depicted in abusive material can have their likeness manipulated into new images. Even when a generated image does not document an actual event, it can still be used to harass, threaten, groom, or blackmail a child.

The UK government has also acknowledged that AI-generated CSAM can have direct effects on real children and can be used in grooming and blackmail.

Are AI-Generated Child Abuse Images Illegal in the UK?

The creation, possession, and distribution of child sexual abuse images are already criminal offences in the UK, including certain AI-generated material, according to Digital Trends. Existing laws cover photographs and pseudo-photographs as well as other prohibited images.

However, lawmakers have identified a gap involving AI models specifically adapted or optimized to produce child sexual abuse material.

The Crime and Policing Act 2026 introduced a new offence covering the creation or adaptation of an AI image generator for producing or facilitating the creation of child sexual abuse images. Possessing, supplying, or offering to supply such a generator can also constitute an offence. The maximum sentence on conviction on indictment is five years’ imprisonment, a fine, or both.

The law is aimed at people who deliberately create, adapt, possess, or distribute tools intended for this purpose. It does not simply criminalize ordinary AI developers because their technology could potentially be misused.

Safety-by-Design Is Becoming a Bigger Issue

The rapid increase in AI-generated CSAM has intensified calls for safety by design.

The basic idea is that safeguards should be incorporated into AI systems before they are released rather than relying primarily on users, law enforcement, or organizations such as the IWF to deal with harmful content afterward.

The IWF has called for stronger legal requirements that would push technology companies to develop AI systems with protections against this type of misuse. Its position is that companies should anticipate foreseeable abuse and build safeguards into their products.

This can include stronger restrictions around image generation, better detection systems, monitoring for attempts to circumvent safeguards, and mechanisms for quickly reporting and removing illegal material.

The challenge is that safeguards must keep pace with rapidly changing technology. A restriction that works against one model or technique may not necessarily stop another system from producing similar content.

Why Parents Are Being Advised to Limit Public Photos

The growing availability of image manipulation has also changed the advice being given to families.

The IWF and the UK’s National Crime Agency have encouraged parents and carers to think carefully about how much identifying imagery of children is publicly available online. Keeping accounts private and limiting children’s photographs to trusted audiences can reduce the amount of material available for misuse.

This does not mean parents are responsible for preventing AI abuse. The responsibility for creating and distributing abusive material remains with offenders.

However, limiting public access to children’s photographs can reduce one potential source of material for image manipulation.

Parents can also review privacy settings on social media accounts, avoid publicly displaying unnecessary identifying details, and talk to children about what they are comfortable sharing online.

What Should a Child Do If an Image Has Been Manipulated?

A child who discovers a fake or manipulated sexual image should not be blamed for what happened.

The safest response is to involve a trusted adult and report the material through the appropriate platform or child-protection service. In the UK, Report Remove allows children under 18 to request help with removing nude or sexual images of themselves. The service also allows them to indicate when they believe an image has been faked or manipulated.

It is also important not to forward or redistribute the image. Sharing it, even when the intention is to show someone what happened, can increase its circulation and potentially create additional legal and safeguarding concerns.

Where threats, blackmail, or suspected criminal activity are involved, law enforcement or a child-protection organization can provide further assistance.

Frequently Asked Questions

Is AI-generated child sexual abuse material illegal?

Yes. In the UK, creating, possessing, and distributing child sexual abuse images, including relevant AI-generated material, is already illegal. The 2026 Crime and Policing Act also introduced a specific offence covering AI image generators that are made or adapted to create such material.

How much AI-generated child abuse material was found in 2026?

The IWF assessed 6,310 AI-generated images meeting the legal definition of child sexual abuse between January and June 2026.

Can AI-generated images harm real children?

Yes. AI-generated or manipulated images can use the likeness of real children and may be used for harassment, grooming, blackmail, or sexual exploitation. UK government guidance specifically recognizes these risks.

What is Report Remove?

Report Remove is a service that helps children under 18 report nude or sexual images of themselves and request their removal. Since 2025, the service has also allowed children to identify imagery they believe has been faked or manipulated.

How can parents reduce the risk of children’s photos being misused?

Parents can limit the public availability of children’s photographs by using private social media accounts and sharing images only with trusted people. The IWF and National Crime Agency have encouraged families to consider these precautions.

Why are AI companies being urged to use safety-by-design measures?

Safety by design aims to prevent harmful uses before they occur. The IWF argues that AI companies should build stronger safeguards into their systems instead of relying mainly on detection and removal after harmful content has already been created.