"What school do you go to?"
"Do your parents check your phone?"
Nothing menacing in either. Both are on a list of examples published on 8 September 2026 by Adelaide University, from new research into how Australian children experience online grooming [1][8].
The headline finding will travel: nearly two-thirds of the Australian children surveyed had received at least one form of grooming-related communication from a stranger online in the past year, and almost one-third reported direct requests for sexual information or interactions [1][2]. The more useful finding is the texture. Most of these children did not feel alarmed. Most believed they were talking to another kid. Most did nothing [7][17].
The study was led by Adelaide University and published in Child Abuse & Neglect, with researchers from eSafety, the Australian Institute of Criminology and Michigan State University [4]. It surveyed 304 children aged 10 to 17, proportionally representing the broader Australian population [16]. The university is careful about how far that stretches: the survey gives "a snapshot of children's online experiences rather than nationally representative estimates" [3].
Within that group, "nearly 66% experienced grooming-related communication, and 31% experienced solicitation" [14]. The survey measured twelve grooming and solicitation strategies, and how often each turned up varied — from 15 per cent for external enticements to 54 per cent for risk assessment [15]. Around one in five children had been asked by someone they only knew online to send photos or videos of themselves, or to meet up in person [6].
Lead researcher Dr Katie Logos said the most common experiences "involved communication related to trust-building, personal questions and risk-assessment behaviours" [5] — shared interests, questions about a child's life and what they do online.
The conversation that ends badly does not open badly.
The release publishes examples of messages that can form part of grooming [8]:
And the warmer ones: comments such as "We have so much in common" or "You're really easy to talk to" [9].
Dr Logos's line is worth keeping: some grooming behaviours "may blend into ordinary online conversations and might be difficult to recognise" [5a]. As the researchers note, children often cannot verify who is behind an account [17].
Dr Logos said the prevalence was higher than expected and "above rates reported in previous studies, which have typically found rates ranging from around 5 per cent to 30 per cent" [12]. Part of that is design: it asked about a broader range of behaviours, and about any online stranger — not only adults.
Then there is where it happens: the research describes grooming often beginning through social media, online games or messaging apps. No platform is named, and we won't name one.
That matters because of what the law reaches. From 10 December 2025, the law requires age-restricted platforms to stop under-16s holding accounts — a set of services that includes Facebook, Instagram, TikTok, Snapchat, YouTube, X, Reddit, Threads, Twitch and Kick, with more having self-assessed since (our guide keeps track). Games are a different question: Fortnite and Minecraft appear on neither eSafety list, and not having been assessed is not the same as eSafety deciding they are exempt.
The chat in a game does not go away because the Instagram account did.
Two findings in the paper sit awkwardly together. Increased digital engagement, "especially posting media, was strongly linked to experiences of grooming and solicitation communication" [18]. But "no link was found with implementing digital safety practices, including parental mediation and privacy measures" [19]. Children using privacy settings, parental guidance or age-appropriate platforms were just as likely to report these messages as those who were not [10].
Professor Russell Brewer, who directs the Australian Cybercrime Observatory, adds the caveat himself: "This does not mean these safety measures are ineffective. It is possible that some children adopted stronger safety practices after experiencing these interactions" [11]. Settings are still worth having. They are just not a wall.
Which is why the researchers land on education rather than configuration: it "may need to go beyond warning children about obvious dangers or direct requests for sexual content" [13]. The obvious dangers are not the ones getting through. "What school do you go to?" is.
This study's hardest finding — that children mostly believed they were talking to another child — is not new to us.
Our oldest son was chatting online with another "child" who wanted him to travel from Brisbane to Sydney to meet up. Luckily we caught it in time.
For the plain-English version of the early stages, our guide to online grooming warning signs walks through them one at a time — because online safety shouldn't stop where the ban does.
Read next: our guide →
Aunty Bea helps Australian families with what comes next, quietly. You'll never scroll your child's messages in Aunty Bea. You get a plain-English summary of what looked risky, and how to talk about it.
Get started free →