Teenagers who use artificially intelligent chatbots built into Instagram, Facebook and Messenger could be asked to explain themselves after Meta announced changes to parental controls.
The US social media giant revealed it would prepare reports about the topics teen users ask Meta AI and share them with parents and guardians, in addition to alerts when teens quiz the technology about self-harm.
The safeguard is one of several announced by Meta to police artificial intelligence on its platforms, and will be introduced to Australia, the US, UK, Canada and Brazil in the coming months.
But the rules also come a month after the eSafety Commission named Facebook and Instagram among five digital platforms on notice for failing to take reasonable steps to prevent children under 16 years from holding an account.
A failure to do so could lead to fines of up to $49.5 million.
Meta announced its changes in a blog post, revealing parents and guardians supervising their children's accounts would receive weekly summaries of the types of questions they ask AI.
"Parents will be able to see the topics their teen has been asking Meta AI about in that specific app over the past week," the company said in a statement.
"Topics can range from school, entertainment and lifestyle to travel, writing, and health and wellbeing."
Teens' exact queries will remain hidden but parents will be able to select categories for greater detail, to see if their child has asked about fitness or mental health within the category of wellbeing, for example.
The weekly subject matter reports will join proactive warnings sent to parents if their teen repeatedly asks Meta AI questions about issues related to self-harm or suicide within a short period of time.
The company also revealed it would create an AI Wellbeing Expert Council with participants from three existing advisory groups to provide recommendations for teens' use of the technology.
The changes come amid heightened scrutiny of the company over its treatment of young users, both in Australia and abroad.
In the US, a judge ordered Meta to pay $US375 million ($A524 million) in March after a jury found it misled users about safety on its platforms and enabled child sexual exploitation.
The company has pledged to appeal the verdict.
Lifeline 13 11 14
Kids Helpline 1800 55 1800 (for people aged 5 to 25)