‘AI-powered manipulation' will still reach children's inboxes under Australia's under 16s ban, experts warn

West Australian

3 days ago

Porn and 'AI-powered manipulation' will still be able to reach
children
's inboxes in Australia's under 16s
social media
ban, as experts warn parents still need to teach digital literacy.
Bubbling beneath the surface of Messenger's chat-interface, users can still access a string of extreme artificial intelligence-powered chat bots and receive unsolicited sexual-related messages.
Unlike YouTube,
Facebook
,
Tik Tok
, X, or
Instagram
, the app won't be excluded under rules released by the Federal Government — despite being owned by Meta, Facebook and Instagram's parent company.
The rules state 'services that have the sole or primary purpose of enabling end‑users to communicate by means of messaging, email, voice calling or video calling' will not be captured in the ban.
While Messenger had previously been part of Facebook's interface, it has become its own separate communication platform — meaning users no longer have to hold an active Facebook profile to access it.
A deep dive into the app has found AI bots based on 'Satan', 'Your drunk dad', 'POTUS Trump', 'Illuminati' and 'Smash or pass' that users can join.
Created by a user called '69.lucifer', the Satan AI bot claims to have sent 1.8 million messages. Once engaged, the bot brags of a world where 'weaklings beg for scraps' and it encourages users to 'speak your failures'.
Another AI bot focused on the violent video game Call of Duty offered up 'tips on a specific gun' and 'weapon builds'.
An AI Bot called 'Smash or Pass?' claims to be 'flirtations and playful' and encourages users to engage in a 'simple game' for rating images of sometimes topless AI-generated characters.
When asked if underage users could play, it sought confirmation the respondent was an adult but swiftly dropped the age-verification request when asked a new question.
Messenger users can also be added to mass chat rooms containing sexually-suggestive messages, pornographic references and external links.
A handful uncovered by The West Australian referenced an 'adult chat room', 'adult webcam chat' and 'passion room'.
La Trobe University senior lecturer Alexia Maddox warned messaging apps had 'weaker safety infrastructure' than major social media sites and called for a government rethink on the ban rather than playing 'whack-a-mole with platforms'.
'These AI chatbots represent a fundamental shift in online risk — they're not human predators we can arrest, they're algorithmic systems designed to exploit psychological vulnerabilities,' she said.
'The social media ban may inadvertently push children toward less regulated messaging platforms where they're actually more vulnerable to these new AI-powered threats.
'Rather than playing whack-a-mole with platforms, we need safety-by-design across the entire digital ecosystem.
'This requires a completely different safety approach than traditional child protection measures.'
As generative AI advances, Dr Maddox expressed concern that 'Satan' and sexual content-type bots would likely only get 'more sophisticated, more persuasive, more psychologically targeted', while legislation failed to keep pace.
Dr Maddox called for greater focus on platform accountability rather than user restrictions. She was sceptical about the effectiveness of a government ban and predicted youth would find their way around it with virtual private networks.
Instead, she flagged the importance of open communication between parents and children about online activities and the need for platforms to create safer spaces.
'We can't ban our way to child safety online. We need platforms to take responsibility for safety by design, and we need to equip young people with the digital literacy skills to navigate whatever spaces they'll inevitably use for social connection.'
She said WhatsApp had already shown via automatic spam detection, easy reporting tools, default privacy settings that other message-based apps could implement greater accountability tools.
eSafety Commissioner Julie Inman Grant had flagged concern during a National Press Club address in June that apps were continuously adapting and adding features, such as AI chat bots.
The Commission has released advice dedicated to parents about the use of AI bots and dangers to young people.
Mr Albanese has acknowledged that the ban won't be foolproof but has publicly argued that it's still worth doing.
Dr Maddox also said there were still unanswered questions about what age verification technology platforms would use.
'Of course, anyone who doesn't have the ability to be recognised by their age tech, or is in some kind of identity transition they will not have access, and that includes all adults,' she added.
'So we definitely need to see what's going to happen with that.'
A spokesman for Communication Minister Anika Wells said while some messaging apps were exempt, platforms still had a 'digital duty of care' to prevent harm under the Online Safety Act.
'The duty will complement existing powers under the Act that allow the eSafety Commissioner to direct the removal of cyberbullying material targeted at a child or cyber abuse material targeted at an adult,' he said.
'We are also creating a safer space online through classifications laws and industry codes.'
The spokesman said the exemption acknowledged messaging apps were an 'important tool' for young people to communicate.