The Thought Police Are Using the Protection of Kids to Justify the Expansion of Surveillance

The Australian federal police is seeking to use AI chatbots to monitor the activity of Australian children online to ensure they’re not being radicalised by terrorist and criminal networks, whilst the Queensland government is set to establish an intelligence hub that will have the ability to share information across state agencies to prevent child sex offenders from working with children.
But as Sisters Inside chief executive Debbie Kilroy put it this week, her organisation is concerned that “children are increasingly becoming the testing ground for forms of surveillance and intelligence gathering that would once have been considered unacceptable in a democratic society”. And she further warns that “child protection should not become synonymous with intelligence gathering”.
As AFP commissioner Krissy Barrett spruiked her department’s initiative, which appears to be linked to the Anglosphere’s global surveillance network the Five Eyes, to the Murdoch press, she initially imagines a chatbot communicating with an Australian child about any criminal plans they might have, before providing a second example involving bots making exchanges with criminal networks.
And in terms of the proposal to create the Queensland Protection Commission to track individuals seeking “blue cards”, or permission to work with children, the Crisafulli government notes that the “intelligence hub” will involve a network that “will break down existing legal barriers that prevent agencies from proactively sharing information with each other”.
Kilroy further states that the “stories being used to justify these proposals are confronting”, and that no one wants children to be exploited or radicalised, but she adds that fear should not be the reason to “undermine privacy, freedom and the rights of children”.
Indeed, citing child sex offenders as reason to expand surveillance laws is an old trick that conservative politicians have long employed.
Government tracking kids online with AI
“Right now, kids are hurting their siblings, killing their pets or injuring themselves because they are being blackmailed by other youth, teens or adults who enjoy watching the pain of others or are dominating and exploiting the vulnerable,” AFP commissioner Barrett told the Sunday Mail on 14 June 2026, as she listed outcomes that the targeting of kids online is causing.
The top cop set out that her concerns are around terror networks radicalising children of between 13 and 16 years old, along with criminal networks recruiting such kids to become gig criminals, or individuals who perform illegal odd jobs. And article intelligence and “echo chamber” algorithms are being employed by these nefarious actors, and that’s why the AFP needs to deploy AI chatbots.
Barrett said last weekend that the Australian Criminal Intelligence Commission will also be in on the caper, and she added that she’d already been in consultation with one tech company and she would also be discussing the initiative further when the Five Eyes Law Enforcement Group met in London over this last week.
The Five Eyes is the 1946-established intelligence sharing alliance that involves the United States, the United Kingdom, Canada, New Zealand and Australia.
The AFP has investigated 63 kids since 2020, with the youngest being 12, and 33 have been charged with Commonwealth offences. A 13-year-old boy living on Gubbi Gubbi and Butchulla land in the Queensland regional city of Maryborough was arrested in May over planning to kill other children at school. And school shootings, bomb making and terror attacks are all key concerns of the AFP.
The article outlining the federal police need to track Australian children online using AI further cites the August 2025 announcement by ASIO that involved received intelligence revealing that two local firebombings on Jewish sites were coordinated by Iranian state actors, and it suggests that the AFP are concerned about foreign state actors targeting local kids as further reason for the initiative.
Using child protection to bust privacy
The Queensland government commissioned the state’s Child Death Review Board to conduct the In Plain Site review of system responses into child sex abuse in December 2024, following former childcare worker Ashley Paul Griffiths having that year been sentenced to life imprisonment over dozens of sexual assaults he perpetrated upon children. And the review report was released last year.
The Crisafulli government has just announced that it will be investing $255 million into the establishment of the Queensland Protection Commission, which will be charged with identifying the risks of child sexual abuse and preventing it before it happens.
The government notes that while child sex abuse criminal cases “appropriately need to be held to a high standard”, the agency “action to protect children should not be reliant on that threshold being met”.
The QPC will be “a nation-leading child safeguarding intelligence hub” with the ability to source intelligence across government agencies in order to “identify those who should not be working with children”. This will involve being able to share information between different government agency databases, and it will “break down existing legal barriers” that prevent this from happening.
“Analytical tools” will be advanced by the QPC, so that rather than assessing individual complaints, the hub will be able to source details on individuals across agencies, and, with its ultimate goal of safeguarding children, the ability to access this information will not be reliant upon whether an individual has any criminal record or has been disciplined in the past.
The Country Liberal’s QPC initiative appears to take its cue from Palantir and the Trump White House in terms of their attempts to eliminate different data silos and integrate them to produce fresh data points. And the issue with Queensland crashing through barriers that prevent private data being shared across agencies, is that once these protections are torn down it won’t just be for the QPC.
Protecting children as a means
“While we recognise the very real harms that can occur online, including exploitation, coercion, grooming, sextortion and violence, expanding surveillance powers over children is not the answer,” Kilroy insisted. “What is being proposed is not prevention. It is the normalisation of mass surveillance and the expansion of policing into the everyday digital lives of children and young people.”
The lawyer explained that “history shows” that when new surveillance powers are legislated in one targeted area, they rarely remain confined to it. She too warns that the breaking down of legal barriers in terms of information sharing will be expanded across the nation’s systems until they become normalised, and this will disproportionately impact communities already overpoliced.
Kilroy adds that the idea that these proposals will prevent crimes before they occur is dubious and rather, predictive approaches have already been shown to “identify vulnerability as risk, poverty as danger and marginalised families as subjects of intervention”. She also warns that the QPC will excessively impact First Nations communities as policing and corrections systems already do.
The National Network of Incarcerated and Formerly Incarcerated Women and Girls member Tabitha Lean warns that in terms of AI being employed to monitor children, the techniques involved are likely to be “expanded and embedded into everyday life”, so what begins as a targeted response to kids will soon become a permanent and all-pervasive feature of the online environment for all.
“Algorithms are not neutral,” Lean underscored. “They reflect the assumptions, biases and priorities of the institutions that create them.”
“We have already seen automated decision-making systems produce discriminatory outcomes across policing, welfare, immigration and criminal legal systems,” the justice advocate added in concluding. “There is no reason to believe that AI-powered surveillance of children will be immune from those same harms.”





