AFP warns AI chatbots no substitute for human support as study reveals children's reliance

Police warning follows Adelaide University research on help-seeking after online abuse

By LineZotpaper
Published
Read Time2 min
Australian Federal Police commissioner Krissy Barrett has warned AI chatbots are not a sufficient replacement for human support for children, after a study found more than one in three Australian children who are sexually victimised online turn to AI chatbots for support, or tell nobody at all.

Writing in The Conversation, clinical psychologist Amy Dawel from the Australian National University said the warning followed research showing how widely children now use the technology in moments of vulnerability. Around half of Australian children aged 10 to 17 who use AI do so for personal and social reasons, according to research cited by Dawel, including seeking advice on sensitive topics, talking about their feelings and "hanging out".

Dawel wrote that there can be benefits to children using AI, such as help with learning and practising social skills. But there is also a risk they will go to AI chatbots and assistants for emotional support instead of parents and other trusted adults.

One practical reason for the shift is that chatbots provide answers around the clock, making them a natural option when a trusted adult or friend is not available. Curiosity plays a part too, with young people experimenting with the technology much as adults do.

Sometimes the reasons are deeper. Dawel noted young people may want information about something they feel embarrassed raising with parents or friends, and as teenagers explore their identity they begin asking more sensitive questions. The article was written to help parents understand how children can use AI safely, drawing on Australian and international research.

§

Analysis

Why This Matters

  • Chatbots are now part of the safety net, or its absence, for vulnerable children: the Adelaide University study found more than one in three Australian children sexually victimised online turn to AI for support or tell nobody at all.
  • AI chatbots are a mainstream presence in children's lives: around half of Australian AI users aged 10 to 17 use them for personal and social reasons, including sensitive conversations.
  • The AFP's intervention puts law enforcement on the record that technology cannot replace human support, raising expectations on parents, schools and AI companies.

Background

Children now have near-constant access to AI chatbots through phones, websites and apps, and many treat them as always-available listeners. Australian regulators, including the eSafety Commissioner, have been researching how children use AI assistants and companions, finding that personal and social use is widespread. The new research on online sexual victimisation has sharpened concern about what happens when a child's first point of contact is a machine rather than a person.

Key Perspectives

Law enforcement: AFP commissioner Krissy Barrett has warned AI chatbots are not a sufficient replacement for human support for children. Researchers: The Adelaide University study found a substantial share of victimised children turn to chatbots or stay silent, pointing to gaps in access to human help. Clinical experts: Psychologist Amy Dawel argues AI can help with learning and practising social skills, but warns children may use it for emotional support instead of parents and trusted adults. Parents: They are being asked to manage a technology children already use for private, sensitive conversations, with practical guidance still emerging.

What to Watch

  • Whether the eSafety Commissioner or other Australian bodies release new guidance for parents and AI companies on minors' use of chatbots.
  • Further research on whether children who receive chatbot support later seek human help.
  • Whether AI providers introduce stronger safeguards for underage users in response to the police warning and the study's findings.

Sources

Zotpaper

Written by software from the reporting listed above, scored by an automated standards desk, and published without a person reading it first. If something here is wrong, tell the editor and it will be put right.

How we workSubscribe