AI-generated child sexual abuse material is proliferating online at a rapidly increasing rate, with new figures from the Internet Watch Foundation (IWF) showing that the number of identified images in the first half of 2026 has already surpassed the total recorded throughout 2025, The Guardian reports. 

Between January 1 and June 30, IWF analysts assessed 6,310 AI-generated images that met the legal definition of child sexual abuse material. That figure is 40% higher than the 4,512 images identified during the entirety of 2025.

The IWF said the material is becoming increasingly realistic, raising concerns over the ability of existing detection and removal systems to keep pace with newly generated content.

The organization also reported that girls accounted for 98% of the images in which both age and gender were recorded during the first half of 2026. Children aged seven to 13 accounted for 79% of the identified imagery.

The IWF said the rise highlights the need for stronger regulation governing AI systems and platforms hosting or distributing illegal material.

Read more: OpenAI's agent hack probe racks up a $500,000 daily bill

Calls grow for stronger AI regulation

The Guardian reports that Hannah Swirsky, head of policy at the IWF, backed calls from parliamentarians for "binding legislation on AI" to compel technology companies to develop safer systems.

"It is incumbent on tech companies to build tools which cannot be abused this way," she said. "This problem has been created by technology, and if companies won’t build tools which are safe by design."

The IWF has also called for stronger European legislation capable of addressing both known and newly generated abuse material.

One challenge is that conventional hash-based detection systems are designed to identify material that has already been assessed and fingerprinted. Newly generated AI material has no existing digital fingerprint and must first be identified before it can be added to detection systems.

UK expands legal measures against AI abuse tools

The UK government has introduced new measures through the Crime and Policing Act 2026 targeting AI systems specifically adapted to generate child sexual abuse material.

Under the legislation, making or adapting such a system, or possessing, supplying, or offering to supply one, can constitute a criminal offense carrying a maximum sentence of five years.

The legislation also strengthens measures concerning online services used to facilitate child sexual abuse and expands existing provisions concerning guidance on producing abusive imagery.

UK law already prohibits creating, possessing or distributing child sexual abuse images, including AI-generated material. The new provisions extend the legal framework to AI systems specifically optimized for producing such content.

AI Minister Kanishka Narayan has previously said "nothing is off the table" regarding further regulation, with the government assessing proposals according to whether they can improve public safety.

AI manipulation creates new risks for children

The spread of AI-generated material also extends to manipulated images involving real children.

The IWF has warned that offenders can use AI to alter genuine images of victims, creating additional material that can further traumatize survivors. It has also reported growing concerns over young people discovering that images of themselves have been manipulated into sexualized material.

The issue has prompted the IWF and the UK's National Crime Agency to advise parents and guardians to exercise caution when sharing photographs of children online.

The IWF has argued that the rapid expansion of AI-generated material demonstrates the need for systems capable of identifying newly created abusive content rather than relying exclusively on databases of previously identified material.

Read more: Developing nations demand greater say in AI governance at UN