Chatbot chatter: crackdown on ‘lawful but awful’ content to protect kids – EducationHQ
Australians using some artificial intelligence chatbots will need to prove their age from March next year in a bid to reduce children’s access to confronting or sexual content online.
Under six regulatory codes released by the commission on Tuesday, platforms which show “high risk” or “harmful” content will be required to check a user’s age before allowing them in.
Under six regulatory codes released by the commission on Tuesday, platforms which show “high risk” or “harmful” content will be required to check a user’s age before allowing them in.
The nation’s digital watchdog is concerned about young people engaging in sexually explicit conversations with AI platforms, sometimes for hours at a time.
The eSafety Commissioner is also hoping to limit children’s exposure to chatbots which have, in some cases, encouraged self-harm or suicidal thoughts.
Under six regulatory codes released by the commission on Tuesday, platforms which show “high risk” or “harmful” content will be required to check a user’s age before allowing them in.
Pornography websites and social media apps which allow pornography, self-harm material or graphic violence are also covered under the rules, which kick in on March 9.
eSafety Commissioner Julie Inman Grant said the policies would help reduce children’s exposure to “lawful but awful” content before they’re old enough to properly understand it.
Lower risk platforms like Copilot or ChatGPT won’t be required to introduce age verification, but will need to have safeguards built in so conversations don’t veer into sexualised content or encourage self-harm.
But chatbots which specifically advertise sexual conversations will need to force users to prove their age.
“We’ve been concerned about these chatbots for a while now and have heard anecdotal reports of children – some as young as 10 years of age – spending up to five hours per day conversing, at times sexually, with AI companions,” Inman Grant said.
“We’ve also seen recent reports of where AI chatbots have allegedly encouraged suicidal ideation and self-harm in conversations with kids with tragic consequences.”
A July 2025 study from researchers at Northeastern University in the United States found large language models like ChatGPT, Gemini and PerplexityAI could be manipulated into sharing harmful content with users, including analysis of different suicide methods.
Separately, a lawsuit lodged in the US alleged ChatGPT contributed to the death of a 16-year-old boy.
Transcripts showed the chatbot offered to help write his suicide note and encouraged him to keep his thoughts a secret from his family.
Peak bodies from across the tech sector which developed the new Australian codes are describing them as a “significant step forward in making the internet safer”.
“It’s vital the rules that apply to the online world keep pace with changes in the threat landscape,” director of industry group Digi Jennifer Duxbury said.
“The new codes are about making sure families have stronger protections, without losing the benefits of digital services.”
AAP
Staff
This story was published by the EducationHQ editorial team.
Kia ora. I have Taught in Auckland and Sydney. Teachers are not given financial incentives and much needed help in …
— Melvin Din on We Still Don’T Know How To Fix The Global Teacher Shortage
Children in frontline areas of Ukraine have lost 25 per cent more days of learning due to the war than the world’s longest COVID-19 school closures, a new study by global charity organisation Save the Children has found.
Enforcement action has been launched by Australia’s eSafety Commissioner against a technology company responsible for AI-generated ‘nudify’ services used to create deepfake pornography of Australian school children.
Early encounters with pornography are playing a major role in shaping attitudes to sex, particularly among boys, and often lead to long-term struggles including addiction and relationship breakdowns.
Education ministers have set the achievement bar too low for Australian students, with current improvement targets at risk of locking in underachievement across some jurisdictions, experts from Grattan Institute have flagged.
Rigorous, independent evaluation is needed so Governments stop delivering ‘solutions’ for recruiting and retaining teachers that don’t work and deliver more of those that do.
Public school teachers, education support staff and school leaders have rallied outside the Geelong electorate office of state Labor Minister, Gayle Tierney, demanding the reversal of what they call the Victorian Government’s ‘$2.4 billion cut’ to…
Many schools are responding to bullying too late in the piece, when unkind and anti-social behaviours are already well established, an expert has flagged.
Personality traits such as conscientiousness and open-mindedness help school children improve their academic performance, new international research has found.
A new machine learning platform for teachers and junior sports coaches to more accurately assess children’s movement skills and to guide improvement, could be a global game-changer in physical education.
Male teachers who are the targets of bullying or harassment in schools may downplay their experiences and lack proper support from leadership and wider systems, research has found.
Australia: +61 3 9229 6200
35-37 Melville Road, Brunswick West VIC 3055 Australia
PO Box 250, Brunswick West VIC 3055 Australia
