.png)
TL;DR
- Bans may reduce exposure, but they will not remove online risk.
- Australia shows the problem: accounts were restricted, but many children still found access.
- Age checks often mean identity checks, with privacy trade-offs.
- Workarounds can make harm harder for adults to spot.
- A single ban is too narrow; safeguarding needs layers.
The intention is right, but the mechanism matters
I do not think anyone serious about safeguarding can dismiss the intent behind these bans. The current online environment is not benign. Children are exposed to content and contact risks that can escalate quickly, and the design of many platforms is built around attention rather than wellbeing.
Australia has already moved in this direction. Its under-16 social media restrictions place responsibility on platforms to take reasonable steps to stop under-16s from creating or keeping accounts, with children and parents not penalised. The UK now appears to be considering a more expansive version, including restrictions on high-risk features such as livestreaming and stranger contact for under-18s.
There is an appeal. It places more responsibility on the platforms that design, monetise and moderate these environments, rather than leaving individual families to manage the issue alone. Parents and schools should not be left to carry the full burden of managing systems they did not build.
A ban is only one safeguarding control. Like any control, it needs to be tested against how people actually behave.
What Australia tells us so far
Australia is useful because it gives us something more concrete than speculation. According to the eSafety Commissioner, millions of under-16 accounts were removed, deactivated or restricted soon after the restrictions came into effect. Regulation can force platforms to act and can change the default position from open access to restricted access.
At the same time, later polling reported by the Molly Rose Foundation suggested that a majority of Australian 12 to 15-year-olds who previously had restricted accounts still had access to at least one. The policy may still be useful, but the evidence makes it more complicated.
Both things can be true at once. The ban may reduce some access, send an important public signal and create pressure on platforms. It may also be unevenly enforced, difficult to measure and relatively easy for some children to work around.
For safeguarding leaders, the important part is whether the policy changes the risk in a way that can be seen, understood and managed.
The workaround problem
Children and teenagers are adaptive. If one route is blocked, many will look for another. They may use an older sibling’s account, change a date of birth, use a VPN, move to a messaging app, create a secondary account or spend more time in smaller online communities that parents and schools are less familiar with.
This is where the safeguarding risk becomes less clear. If a child is using a large mainstream platform, there may at least be some moderation, reporting tools, parental controls and regulatory scrutiny. Those systems are imperfect, often very imperfect, but they do exist.
If the same child moves somewhere parents, schools and platforms understand poorly, the adult world may have fewer signals to work with. Harmful contact, grooming behaviour, bullying, extremist influence, sexualised content and coercion do not need a mainstream social platform to exist. They need access, secrecy and a channel.
A ban may reduce visible participation on certain platforms while increasing hidden participation elsewhere. If that happens, the safeguarding environment may become harder rather than easier to understand, which is not a great outcome.
Age assurance is really a digital identity debate
The practical question behind any under-16 ban is age assurance. How does a platform know that a user is 15, 16 or 17?
The answer usually involves some form of evidence: government ID, digital ID, facial age estimation, bank-account checks, voice recognition, behavioural inference or another verification method. Each option creates trade-offs. Stronger checks may be more reliable, but they also raise concerns about privacy, exclusion, data security and the normalisation of identity checks for everyday online activity.
This is where the public debate tends to become too narrow. It is described as a social media ban, but it also raises questions about digital identity infrastructure.
Who verifies age?
Who holds the data?
What happens when a child does not have the required documents?
What happens when the system is wrong?
What safeguards prevent identity data being reused for purposes beyond child safety?
These are governance questions, not just technical questions. They should be treated with the same seriousness as any other control that affects children, privacy and access to public digital spaces.
A ban may work better as a signal than as a wall
Under-16 social media bans may be more useful as a public awareness signal than as an impenetrable wall.
A wall suggests that children below a certain age are kept out. In reality, many will still find a route in. A signal does something different. It tells platforms, parents, schools and children that the environment carries risk and that access should not be treated as harmless by default.
Public health warnings do not stop every harmful behaviour, but they can change expectations, place responsibility more clearly and make the risk harder to ignore. A social media ban may do something similar if it prompts better platform design, more serious parental conversations, stronger school guidance and clearer expectations around high-risk features.
The signal is only useful if it leads to practical action. A warning without education can be unhelpful. A restriction without trusted reporting routes may make children less likely to ask for help. A platform duty without transparency is difficult to trust. A technical control without human judgement will miss context.
What a layered approach should look like
The better answer is layered safeguarding, rather than relying on either blanket restriction or no restriction at all.
First, platforms should be required to reduce harmful design and high-risk functionality for children.
Stronger controls are needed around stranger contact, livestreaming, algorithmic amplification, direct messaging, age-inappropriate content and recommender systems that pull children into escalating risk.
Second, age assurance should be proportionate.
The level of verification should match the level of risk. A passive content feed, a private messaging feature and a livestreaming tool do not carry the same safeguarding profile, so they should not necessarily require the same control.
Third, schools and parents need education that reflects how children actually behave online.
Telling a child not to use a platform is unlikely to be enough. They need to understand grooming tactics, coercion, harmful communities, identity misuse, sexual exploitation, addictive design and how to ask for help without fear that the first response will be punishment.
Fourth, reporting routes need to be trusted and usable.
If a child has bypassed an age restriction and then experiences harm, the system must still make it easy for them to report. Otherwise, the rule may become a reason to stay silent.
Finally, there needs to be human judgement.
Technology can enforce, flag and organise.
It cannot understand the full context of a child’s welfare, family situation, school environment or vulnerability. People still need to decide what is relevant, fair and proportionate.
The broader safeguarding lesson
The under-16 social media debate is really a case study in modern digital risk. Traditional controls often assume that risk appears in known places: a platform, an account, a formal report, a device or a recorded incident. Digital risk does not always behave like that. It moves, fragments and hides. It often becomes visible through patterns and behaviour before it appears in any formal system.
Visibility matters because safeguarding teams cannot assess what they cannot see. Without a clear view of emerging risk, they cannot respond proportionately or explain later why a particular decision was reasonable.
Under-16 social media bans may form part of the answer. They may reduce some exposure, force platforms to take more responsibility and help society stop pretending that the current online environment is fine for children.
On their own, though, they are not enough. The real test is whether they improve our ability to protect children, or whether they push risk into places where children are still exposed and adults are less able to see what is happening.
Let us keep coming back to that question.
If your organisation is reviewing digital risk in a safeguarding context, Safehire.ai can help. Our Digital Risk Screening combines AI-scale discovery with human analyst validation, giving teams clearer and more defensible visibility where traditional checks and informal searches leave gaps.
.png)
.png)
