Protecting children online is not an abstract policy debate for me. It is one of the responsibilities I take most seriously as a journalist, publisher and member of the community.
Public discussion often concentrates on predators, grooming, sexual exploitation and adults who deliberately manipulate children. Those dangers are real, severe and deserving of relentless scrutiny.
But, now here comes the problems I have personally been observing in plain sight, they are not the only threats reaching children through their phones.
A young person does not need to be contacted by a criminal to be harmed online. Sometimes the damaging voice belongs to an influencer they admire, trust and watch every day.
The content may look harmless. It can arrive as a fitness routine, relationship lesson, mental-health confession, beauty tutorial, dietary plan or motivational speech.
Yet popularity is not a qualification. Confidence is not evidence, and a large following does not make someone competent to advise children about health, food, exercise or emotional distress.
That distinction matters because children are not merely searching for information online. Platforms are selecting it for them, repeating it and pushing it into their feeds.
Harm Is Often Delivered by the Algorithm
Fresh research released by Australia’s eSafety Commissioner in 2026 shows just how deeply online platforms have entered childhood.
A nationally representative survey found 96% of Australian children aged 10 to 17 used social media, while 95% used online communication platforms.
The same research found 74% had encountered content associated with harm, including dangerous challenges, misogynistic material and content encouraging unhealthy eating or exercise habits.
This was not limited to a small fringe of highly active users.
Forty-four per cent of children had seen material encouraging unhealthy eating or exercise. Another 19% had encountered content suggesting how someone could deliberately hurt or kill themselves.
Almost seven in ten children said their most recent exposure happened accidentally, often through a social media feed or pop-up rather than a deliberate search.
That finding cuts through one of the technology industry’s most convenient arguments: that users simply receive the content they choose to view.
Children cannot meaningfully choose material they did not seek, delivered by systems they cannot inspect and shaped by commercial incentives they do not understand.
YouTube accounted for 32% of children’s most recent exposure to content associated with harm, followed by TikTok at 24%, Facebook at 12% and Instagram at 11%.
Those social media platforms aren’t passive noticeboards. Their recommender systems decide which voices receive attention, which claims spread and which anxieties are repeatedly placed before vulnerable users.
In a May 2026 position paper, eSafety warned that recommender systems can expand the reach of harmful material while features such as autoplay and infinite scrolling encourage excessive engagement.
The business model rewards attention. Outrage, fear, envy and insecurity can hold that attention far more effectively than careful, qualified and balanced advice.
The Influencer Does Not Need Malicious Intent
Some influencers knowingly exploit their audiences. Others may genuinely believe the advice they provide, despite having no training or reliable evidence to support it.
The outcome for a child can still be harmful.
An influencer may present extreme food restriction as discipline, compulsive exercise as commitment or dangerous challenges as entertainment.
Others may give simplistic advice about depression, anxiety, relationships or identity while dismissing professional support as weakness, corruption or unnecessary interference.
Children may also be encouraged to distrust parents, teachers, doctors or institutions whenever those adults challenge the influencer’s claims.
This is what makes seemingly innocent misinformation so difficult to confront. It is often wrapped in humour, friendship, aspiration and personal storytelling.
The creator does not appear threatening. They appear relatable.
They speak directly into a child’s bedroom, sometimes several times a day, building familiarity that no traditional advertisement could easily reproduce.
Repeated exposure can make an unsupported claim feel normal. A sponsored opinion can look like personal advice, while an algorithmic recommendation can be mistaken for social proof.
Cyberbullying Remains Widespread
The more visible forms of online abuse have not disappeared while these quieter dangers expand.
eSafety’s 2026 findings show 53% of Australian children aged 10 to 17 had experienced cyberbullying at some point, while 38% had experienced it during the previous year.
The burden was even higher among vulnerable groups. Cyberbullying had affected 81% of trans and gender-diverse children, 67% of First Nations children and 63% of children with disability.
These statistics represent humiliation, exclusion, fear and distress entering homes through devices that children also need for learning, friendships and ordinary participation in modern life.
The solution cannot be to tell young people simply to switch off, toughen up or make better choices.
Adults built these platforms. Corporations operate them. Governments regulate them, and advertisers fund the systems that determine which content becomes profitable.
Responsibility must follow power.
Australia’s Age Restriction Is Not Enough on Its Own
Australia’s social media minimum-age law came into effect on 10 December 2025, requiring covered platforms to take reasonable steps to prevent children under 16 from holding accounts.
The law may delay some children from joining major platforms, but it cannot be treated as proof that the wider safety problem has been solved.
An observational study of 408 Australian adolescents found about 85% of under-16 participants were still using social media three months after the law took effect.
The early research found limited compliance and significant circumvention, although its size and timing mean it should be treated as an initial assessment rather than a final verdict on the policy.
The result still carries a clear warning: legislation means little when platforms can satisfy regulators with weak checks while children continue entering the same algorithmic environment.
Removing an account is not the same as removing the danger. Harmful material can still circulate through shared devices, messaging services, embedded videos, search results and unrestricted platforms.
Australia needs age safeguards, but it also needs safer systems.
Parents Cannot Carry This Alone
Parents are routinely told to supervise accounts, check devices, set limits, learn every new platform and identify every emerging threat.
Most are trying.
eSafety found 92% of parents had online-safety rules and 84% held open conversations with their children. Yet 58% still said they struggled to keep them safe online.
Only 26% discussed algorithmic influence, while just 11% talked about how generative artificial intelligence could be used or misused.
That is not evidence of parental failure. It shows how unreasonable the burden has become.
No family can manually audit billions of recommendations, hidden commercial arrangements, changing platform designs and influencers appearing across multiple services.
Parents need support, accessible digital literacy and clear reporting pathways. Schools also need the resources to teach children how online authority is manufactured.
Children should learn to ask who made a claim, what qualifications they hold, what evidence supports it and whether money, sponsorship or attention may be influencing the message.
Mental Health Requires Careful Reporting
Australia’s broader youth mental-health figures make careless online advice particularly concerning, but the evidence must not be overstated.
The Australian Institute of Health and Welfare’s Australia’s Health 2026 report found the proportion of people aged 16 to 24 experiencing a mental disorder rose from 26% in 2007 to 39% in 2020–2022.
That does not prove social media caused the increase. Youth mental health is shaped by family circumstances, economic pressure, education, housing, discrimination and many other factors.
But it does show why unqualified mental-health content aimed at young people cannot be dismissed as harmless speech or entertainment.
When a child is struggling, bad advice can delay real support. It can intensify shame, encourage secrecy or falsely present a complex condition as something cured through discipline, products or positive thinking.
Platforms and Influencers Must Be Accountable
Protecting children online requires a wider definition of harm.
It must include grooming, sexual exploitation, stalking, sextortion and cyberbullying. It must also include algorithmically amplified misinformation that damages health, confidence or judgement.
Platforms should be required to assess risks before releasing features, provide independent researchers with meaningful access and explain why particular material is being recommended to children.
Influencers discussing health, nutrition, exercise or mental wellbeing should clearly disclose relevant qualifications, sponsorships and commercial interests.
Repeatedly distributing dangerous falsehoods to young audiences should attract consequences, even when the content is presented as lifestyle advice rather than an explicit instruction.
Regulators must also be prepared to investigate the systems promoting harmful material, not merely the individual posts visible after damage has occurred.
Children Deserve Better, Need Better Than a Warning Label
I care deeply about this issue because children deserve the freedom to learn, communicate and explore online without being treated as data points, advertising targets or captive audiences.
Protecting them does not mean denying every benefit of technology. Online communities can provide education, friendship, creativity and vital support.
But those benefits cannot become an excuse for corporations to minimise foreseeable harm or shift responsibility onto children and exhausted parents.
The most dangerous person online may be a predator hiding behind a false identity. It may also be a familiar face delivering false advice to millions while an algorithm quietly expands their reach.
One threat looks frightening. The other looks friendly. Our children need guidence and protection from both.
Online, I have seen some parents openly admit helping their children bypass Australia’s new social media age restrictions because they trust their judgement and believe they are mature enough to cope online.
But the issue is not whether parents trust their children. A sensible child can still be manipulated by algorithms, misleading influencers and confident people presenting opinion, sponsorship or misinformation as fact.
The real question is whether these platforms can be trusted to protect young users. Their systems are built to hold attention, not to ensure every message reaching a child is safe, accurate or responsible.
