Research reveals the frequency of AI in online sexual victimisation of Australian children, and how often young people are turning to AI for help.
The study of almost 2000 Australians aged 16 to 18 was conducted in early 2026 by Adelaide University’s Australian Cybercrime Observatory, and is the first national study to measure the role of AI in online child sexual victimisation at a population level.
The research found:
- Nationally, more than one in every three adolescents asked an AI for personal life advice each week (36.6%), while more than one in five asked for psychosocial counselling each week (22.2%).
- At least one in 25 Australians have experienced, or have a close friend who has experienced, online sexual victimisation involving AI before the age of 18. AI is involved in more than one in four cases (26%) of nonconsensual sharing of sexual images.
- Girls are victimised at higher rates than boys; yet when AI is involved, boys are significantly more likely to be targeted.
- Children now disclose experiences of online sexual abuse more to AI chatbots than to authorities: 18.7% told an AI, compared with 13.2% who told a teacher, doctor, counsellor, police officer or helpline.
- More than a third of victims told no person at all, disclosing only to AI or to no one.
- Victimisation clusters within friendship groups, meaning those who had a friend experience victimisation were more likely to experience it themselves.
- Nationally, more than one in every three adolescents were worried that AI would be used to make a fake image of them, or someone they care about.
The Australian Cybercrime Observatory conducted the research in collaboration with the International Centre for Missing and Exploited Children Australia (ICMEC), and the Australian Federal Police-led Australian Centre to Counter Child Exploitation. The research was published in Child Abuse & Neglect, with a summary report available at the ICMEC website.
"There's been a seismic shift in how young people experience and seek help for online sexual victimisation,” said lead author Associate Professor Timothy Cubitt from Adelaide University’s Australian Cybercrime Observatory.
“AI involvement in this abuse is worryingly common, and victims are now more likely to tell an AI than teachers, doctors, police or reporting services. We must adapt quickly to protect children."
The findings point to three priorities: reporting pathways that meet children where they already turn, including AI tools; prevention built for peer networks, given harm clusters within friendship groups; and prevention that reaches beyond traditional risk groups, given the disproportionate impact on boys when AI is involved.
"This research confirms what frontline services and law enforcement have been warning about. AI is now shaping how children are harmed, and how they reach for help,” said Colm Gannon, Chief Executive Officer of ICMEC Australia.
“The reforms of the past year matter, but the response has to move as fast as the technology. That means industry, government and services working together, not in isolation."
Media contacts:
Associate Professor Timothy Cubitt, Australian Cybercrime Observatory, Adelaide University. Email: tim.cubitt@adelaide.edu.au
Lara Pacillo, Media Officer, Adelaide University. Mobile: +61 403 659 154. Email: lara.pacillo@adelaide.edu.au
Dannielle Kelly, Head of Government Affairs, Law Enforcement Outreach, and Communications, ICMEC Australia. Email: press@icmec.org.au