New research from the University of Adelaide has found that around one student in every Year 12 classroom in South Australia has experienced or knows someone who has experienced online sexual victimization involving artificial intelligence.
This could include generating AI images, using AI to alter photos that young people post online through “nudification” apps, and sharing these images without their consent.
A University of Adelaide study released today found that AI was used in one in four cases of image-based child sexual abuse.
Associate Professor Tim Cubitt from the University of Adelaide said this was an “underestimation” by cautious researchers.
This comes after a Mercedes University graduate became the first person in Australia to be charged under federal deepfakes laws in April, pleading guilty to creating and distributing explicit deepfake images.
The 19-year-old became the first Australian to be charged under the federal offense, created in 2024 to combat the creation of explicit images using AI deepfakes.
SA Police said indaily It was “aware that artificial intelligence (AI) apps were being used to generate illegal and false child exploitation images within the community.”
“Police encourage all parents and guardians to talk to their children about online safety. If they would like information to help with these discussions, Police encourage parents to visit the ThinkUknow website,” a police spokesperson said.
Mr Cubitt said the new research showed a “really seismic shift” in disclosure, with AI chatbots being used not only to create exploitative material but also as a tool for young people seeking help.
Of the young people surveyed, 18.7% said they had shared their experiences with AI, compared to 13.2% who had told teachers, doctors, MPs, police officers or helplines.
“This is a very significant change and it means we have to adapt to it… It’s very important that we are comfortable with these platforms and that we like using them,” Cubitt said.
“If you can imagine someone under the age of 18 looking at their phone, they may find themselves potentially victimized by AI-powered image manipulation.
“Probably the quickest advice is to flick into the ChatGPT app or flick to Claude and say, ‘This happened to me. What should I do? Well, that pipeline is really important.'”
The study was conducted in early 2026 by researchers from the University of Adelaide, the Australian Federal Police’s Australian Center for Combating Child Exploitation and the International Center for Missing and Exploited Children (ICMEC), which surveyed approximately 1,900 Australians aged 16 to 18.
More than two-thirds of those surveyed asked AI for advice on their personal lives, and one in six under-18s said they asked AI for personal or psychosocial help at least once a week.
Cubitt said there is currently no legal requirement for AI companies such as OpenAI, which runs ChatGPT and Anthropic behind Claude, to redirect child abuse queries.
“There is an ongoing discussion in Australia regarding digital duty of care for online platforms,” Mr Cubitt said.
“There are big concerns about whether they’re receiving the right information. And I don’t think it’s enough to just say, ‘Here’s a phone number to call,’ to absolve them of their responsibility to provide information to young people.”
“They’re looking for help, they’re looking for comfort, and they’re going to disclose really toxic and harmful personal experiences to these platforms.
“I think we all have a duty to make sure they get the right advice and find the right help.”
The scale of young people’s use of AI will be at the top of the agenda at ICMEC’s national SaferAI for Children Coalition forum, which will bring together representatives from not-for-profits, education and child protection in Sydney on July 8.
Ahead of the event, Sonya Ryan, CEO of the Carly Ryan Foundation, said AI was a “national issue” that exceeded the response of the child protection sector.
Ryan founded a foundation to fight online predators in 2010 after Carly’s death in 2007. Carly was 15 years old when she was abused and drowned on a Port Elliot beach by a 50-year-old man who lured her to a Port Elliot beach after 18 months as an 18-year-old musician.
“Our department has seen AI reshape how children experience harm and seek help, and we cannot respond to that change by working in silos,” Ryan said.
“This will require industry, government, law enforcement, and the services that children actually use to work together, at the same pace as technology.”
If you or someone you know has been affected by child sexual abuse or online exploitation, there are support services available online at the Australian Anti-Child Exploitation Center.
SA Police said members of the public with information about people involved in child abuse or exploitation are urged to contact Crime Stoppers online or remain anonymous on 1800 333 000.
