Latest
AI Affairs, home

Monday 16 November 2026

Government

Childlight reports 8,400% rise in AI child abuse material reports

Unveiled at a ministerial summit in the Philippines, the Into the Light Index calls for better government measurement and earlier prevention of child sexual exploitation.

University of Edinburgh’s New College courtyard, with stone towers and central steps
Photo: Old College Quad Colorized.png : Jesús Quiles Derivate ive work: Hogweard, Public domain, via Wikimedia Commons

Childlight – Global Child Safety Institute unveiled its 2026 Into the Light Index at a ministerial summit in the Philippines on 16 November, calling on governments to treat child sexual exploitation as a preventable public health emergency. Reports of AI-generated child sexual abuse material sent by technology companies to the US-based National Center for Missing and Exploited Children rose from about 4,700 in 2023 to more than 400,000 in 2025, the Guardian reported. Representatives of more than 100 governments attended the summit, the Belfast Telegraph reported.

Key points

  • Reports of AI-generated child sexual abuse material to the National Center for Missing and Exploited Children rose by more than 8,400% between 2023 and 2025.
  • Childlight cautioned that improved detection and reporting may account for some of the increase.
  • Only 22 of 103 countries and areas examined had representative survey data covering both offline and technology-facilitated abuse.
  • Childlight wants governments to invest in consistent measurement, strengthen prevention and hold the technology sector more accountable.

More than 400,000 reports in 2025

Childlight is a research organisation hosted at the University of Edinburgh. Its index examines sexual exploitation both online and offline, with this edition concentrating on east Asia and the Pacific, Latin America and the Caribbean, and North America. The AI figure counts material reported by technology companies to the National Center for Missing and Exploited Children, an organisation that receives reports of suspected child sexual abuse material. Childlight put the rise in those reports at more than 8,400% over two years.

The material includes deepfakes made with apps that place a real child’s face into a sexually explicit image. Researchers said such images can be used to blackmail children for money or further images. The reported increase does not measure only a change in the amount of material being produced: Childlight cautioned that changes in detection and reporting may have contributed to it. Debi Fry, Childlight’s global director of data, said better detection and reporting may explain some of the rise and urged governments and companies to understand emerging forms of abuse, prevent them and remove material rapidly.

There is a separate measure of material assessed by the Internet Watch Foundation, a UK-based organisation that monitors child sexual abuse material globally. Its analysts assessed 6,310 AI-generated images meeting the legal definition of child sexual abuse in the first half of 2026, 40% more than the total of more than 4,500 for all of 2025, the Guardian reported. Those are images assessed by the foundation, rather than reports passed by companies to the US organisation. AI Affairs has also reported on a South Korean AI system that flagged 96,210 child sexual abuse items in six months.

Childlight seeks data across 103 countries

Childlight estimates that more than 300 million children worldwide experience at least one form of technology-facilitated sexual exploitation or abuse each year. It also estimates that about one in four children experience at least one form of abuse before turning 18. The annual estimate concerns technology-facilitated abuse; the childhood estimate covers abuse more broadly. Childlight’s research includes offline sexual abuse, alongside the reports concerning AI-generated material.

Among 103 countries and areas examined across the three regions, only 22 had representative surveys covering both offline and technology-facilitated abuse, the Belfast Telegraph reported. Childlight said children with disabilities, Indigenous children, migrants, very young children, children in humanitarian settings and those abused within their families were particularly underrepresented. Its call for better measurement concerns those gaps as well as the rise in reports sent to the National Center for Missing and Exploited Children.

Childlight also reported that almost a third of girls and more than one in 10 boys in North America had experienced in-person sexual abuse during childhood. In North America, it found that between 2024 and 2025 the frequency of reports and notifications involving child sexual abuse material rose by 63% in the US and by 18% in Canada. Those figures concern reports and notices involving child sexual abuse material generally, rather than the narrower count of AI-generated material sent to the US organisation.

Stanfield calls for a public health response

Paul Stanfield, Childlight’s chief executive, called for governments to recognise child sexual exploitation and abuse as a preventable public health emergency. He argued that governments need better evidence about where harm occurs and which children face the greatest risk if they are to prevent abuse and support children affected by it. Childlight is urging governments to invest in reliable, consistent measurement, use that evidence to strengthen protection and support survivors, and begin prevention earlier in childhood.

The institute also wants greater accountability for the technology sector and laws that keep pace with emerging threats, according to the Belfast Telegraph. Childlight reported that several countries have strengthened online safety legislation, improved detection and put more emphasis on prevention. Its appeal at the summit asks governments to use evidence gathered across surveys, police records and other sources to direct their protection efforts.

Sources

Topics: Regulation, Safety

What should we cover? Tell us

The weekly briefing

The AI Affairs weekly briefing: the week's 10 most important AI stories in business, government and markets, every Monday. We'll email you a link to confirm. Every email has a one-click unsubscribe. Privacy policy