Boys are main victims of online abuse helped by AI

The topic is AI.

It is Tuesday, July 7, 2026, at 4:46 am.

A study found that about one student in every year 12 classroom has been targeted with non-consensual sexual images made using AI.

In short:

A study of almost 2,000 Australians aged 16 to 18 found that one in 25 experienced or had a friend who experienced AI-assisted online child sexual abuse.

AI was involved in one in four cases of non-consensual sexual image sharing, and young people were more likely to tell AI about abuse than authorities or support services.

What's next?

Experts say AI companies need better systems to help those who report abuse and send them to the right services.

About one in every 25 Australians under 18 has been a victim or has a friend who has been a victim of AI-assisted online sexual abuse.

This number is like one student in every year 12 classroom and comes from the first big study on AI's role in online child sexual abuse in Australia.

The study shows AI is both a tool for abusing young people and a way for them to get help after abuse.

Teenagers are using AI chatbots like ChatGPT or Claude to report or get help for online sexual abuse more than they tell police or other authorities.

Support services:

Tim Cubitt from the Australian Cybercrime Observatory said the findings show AI has become a big part of online abuse.

He said this is a big change in how we understand young people's vulnerability and victimization online.

This is the first big study on child AI-assisted abuse in Australia.

The study, done with the International Centre for Missing and Exploited Children Australia and the Australian Federal Police, surveyed 1,894 Australians aged 16 to 18.

Researchers asked about experiences with non-consensual sexual image sharing and solicitation by adults.

They found 15% of young Australians had experienced online child sexual abuse, including non-consensual image sharing.

Is it possible to make a safe AI therapy bot? Researchers tested an Australian-designed chatbot.

About one in four of those cases involved AI.

This often involved deepfake or nudify services that use AI to make explicit images from real photos.

Professor Cubitt said these numbers are likely the minimum because some respondents were unsure if AI was involved.

The study also found that when one young person experienced online abuse, others in their friendship group were more likely to have had the same experience.

Boys are major victims of AI-assisted abuse.

The study found 19% of girls and 11% of boys experienced online sexual abuse, which matches previous research.

But with AI, boys are more likely to be victims.

A larger proportion of online abuse against boys involved AI, with 27% of boys' experiences involving AI compared to 9% of girls.

Professor Cubitt said this is a significant change from past patterns of online abuse.

This is something new in online abuse.

Boys were more likely to report AI-generated abuse.

Boys were more likely to seek help from AI, whether they were victims or not.

Most young people who experienced online abuse told someone, usually friends or parents.

The next most common people they told were AI chatbots like ChatGPT or Claude.

This was more common than telling authorities like police or teachers.

Many respondents asked AI for counseling or told AI about hurtful experiences each week.

The head of ICMEC Australia said AI should be developed ethically so chatbots can refer people to services after disclosures.

The goal is to get them to services that can give them a warm welcome.

How AI companies are responding:

While AI companies have safeguards, there's little information on how they handle disclosures or help-seeking around abuse.

OpenAI and Anthropic didn't respond to questions.

OpenAI's policies prohibit using its services for abuse or exploitation.

It bans and reports users who try to generate or upload child abuse material.

Anthropic also prohibits child abuse material and reports it.

OpenAI's ChatGPT has a minimum age of 13, while Anthropic's Claude is for those 18 and older.

The study found AI chatbots are the most common source of support for AI-generated abuse victims.

New rules for AI providers require services to have reporting systems, but don't mandate reporting disclosures.

Professor Cubitt said the issue is whether young people who disclose abuse to AI get proper advice and referrals.

More disclosures could help young people if they're made to sources that can support them.

There's a duty of care among tech companies and AI platforms.

The next step is to ensure disclosures are supported and referred to the right services.

It is Tuesday, July 7, 2026, at 4:46 am.

It is Tuesday, July 7, 2026, at 8:16 am.

See more ABC coverage.

Pratique d'écoute

Pratique d'écoute