Social Media Restrictions for under 16
Bad idea. I think it is better to educate and to empower
Australia is preparing to block everyone under sixteen from using social media. It is a historic move. It is also a step in the wrong direction.
The intention is understandable. Parents are worried about mental health, sextortion, bullying, and predatory behaviour online. The instinct is to draw a hard line, build a wall, and keep children out. But when you look closely at how young people live, how predators operate, and what technology can actually do today, the restriction is likely to miss the mark.
The existence of social media does not harm children. They are harmed by bad actors who go online. When the environment is unsafe and when adults with harmful intent are allowed to move freely, that’s where the danger is.
We already have the tools to detect those adults early. We can build safer environments. We just haven’t deployed these tools at scale.
Instead of restricting young people from social media, we should be using AI and thoughtful design to remove the danger before it reaches them.
When You Push Teenagers Out, They Do Not Disappear
Keeping teenagers from Instagram or TikTok does not remove their desire to connect. They will not return to an offline childhood because the law says so. They will just shift into places that are harder for parents, educators, and safety teams to see.
They will move to:
Discord servers
Encrypted group chats
Gaming voice channels
Anonymous message boards
Decentralised platforms with no central authority
These are the places where predators operate easily. These spaces have weaker moderation and fewer safeguards. The ban pushes young people into the dark corners of the internet, away from the well-monitored platforms.
It is the equivalent of closing the playground and leaving only the abandoned lot open. The child does not stop playing. The environment just becomes more dangerous.
The Ban Overlooks How Young People Actually Live
We often think of social media as a source of distraction. Endless scrolling and mindless media consumption. For many teenagers, social media is also a source of connection and community.
Consider a teenager in a remote town who has never met another person who shares their identity. Consider a neurodivergent student who finally finds peers who understand how they communicate. Or a child struggling in a difficult home who relies on online friendships to get through the week.
For these young people, social media is not just entertainment. It is connection, community, and sometimes the only available support system. Removing it without offering an alternative creates an emotional void.
Even mental health research shows that moderate social media use can support wellbeing. Total disconnection increases loneliness, which is a serious risk factor for young people.
A blanket ban disrupts the very thing that helps many teenagers stay afloat.
A Ban That Compromises Privacy for Adults and Still Fails to Stop Kids
To enforce an age limit, platforms must verify every user. That means identity checks for adults who have never harmed anyone. You cannot enter without proving who you are.
This has serious consequences for privacy.
And yet it will not stop teenagers.
Install a VPN, press one button, and appear online as if you live in another country. Modern VPNs blend in with ordinary traffic, so platforms cannot reliably detect them without invasive network monitoring.
Teenagers will learn how to bypass school Wi-Fi restrictions. Parents lose privacy. Children bypass the system. And predators continue to operate.
This is not a feasible solution. It places a burden on everyone except those who intend to harm.
We Already Have the Technology to Stop Predators
What we need is not a wall around the internet but shields inside it.
AI can now detect grooming and predatory behaviour early, accurately, and without spying on private conversations. These techniques exist, and they work. They have been used in child safety research for years and are improving rapidly.
The focus should not be on blocking the child. It should be on identifying the adult who intends to cause harm.
AI Can Spot Grooming Patterns Before the Predator Escalates
Grooming is not a random act. It follows predictable patterns. AI can detect these patterns by focusing on shifts in tone, timing, and behaviour.
For example:
An adult messaging a teenager at high frequency
Attempts to move the conversation to a different app
Flattery mixed with secrecy
Gentle pressure that becomes more explicit over time
These are patterns that humans often miss, but AI can identify across millions of conversations. When a risk is detected, the system can block the adult, freeze the chat, or alert the young user. It can intervene long before a predator gains control.
This is how protection should work.
AI Can Protect Privacy with On-Device Safety Tools
One of the strongest arguments for this approach is that it does not require platforms to read messages at all.
With on-device models, the safety system runs locally on the phone. Messages never leave the device. The AI looks for harmful patterns inside the phone itself, then sends only a safety signal if necessary. It protects privacy by design.
Imagine a child typing a message and receiving a quiet warning:
“This person is asking for private photos. Please stop and speak to an adult.”
“This message looks threatening. Do you want to block this user?”
No one else needs to see the conversation. The intervention is instant and private.
This eliminates the tension between child protection and data privacy.
Encrypted Data Can Be Scanned Without Being Decrypted
Another breakthrough is the ability to scan encrypted content in the cloud without opening it. The data stays locked. The AI never sees it in plain form. Yet it can still flag harmful material or known abuse signatures.
This is possible with emerging encryption techniques that allow computation on scrambled data. It preserves privacy while improving safety. It dissolves the old claim that we must weaken encryption to protect children.
We no longer have to choose between safety and privacy.
AI Can Identify Adults Pretending To Be Teenagers
Predators often claim to be younger than they are, but their behaviour usually gives them away.
Adults type at different speeds. They use more complex language. They interact with the platform differently. They move through a conversation in ways that teenagers rarely do.
AI can recognise these patterns. If an account claims to be a child but behaves like a forty-year-old, it can be flagged, restricted, or removed before any harm begins.
This is a far more effective strategy than trying to keep children out entirely.
Fix the Design, Not the Access
If we want safer online spaces, we should change how these platforms work for young users.
Good design can reduce the most common sources of harm. For example:
Private accounts by default
Blocking unknown adults from contacting minors
No algorithmic feeds for under sixteen
No autoplay or infinite scroll
Notifications paused during sleep hours
These settings make the space calmer, slower, and far less attractive to predators. They reduce compulsive usage without removing the platform altogether.
Parents need help building safe environments, not a law that seeks to remove them entirely.
Teach Children How to Navigate the Internet, Not Fear It
The safest teenagers are not the ones who have never touched the internet. The safest teenagers are the ones who know what danger looks like and how to protect themselves.
A more holistic approach to online safety could include:
A Digital Driver’s License for under sixteen
Practical modules on privacy, grooming, scams, and healthy use
Peer-led programs where older students teach younger ones
Regular digital citizenship lessons woven into the school curriculum
Young people already live online. They need digital skills as much as they need literacy and numeracy.
Education creates resilience, while bans create blind spots.
Cut Off the Criminal Networks That Target Kids
Sextortion is run by organised groups. They rely on fast payments from frightened teenagers. We can disrupt these networks directly.
Banks can use AI to detect suspicious payment patterns from youth accounts. Law enforcement can monitor the financial infrastructure behind these crimes. When you choke the revenue of these syndicates, the exploitation drops sharply.
This approach targets the real cause of harm instead of restricting every teenager who just wants to talk to their friends.
Australia Needs a Better Path Forward
The ban aims for safety but creates a long list of unintended consequences. It pushes children into hidden online spaces, removes vital support networks, weakens adults’ privacy, and still does not stop teenagers from logging in through simple workarounds.
The better solution is already available. Use AI to detect danger. Use design to limit harmful features. Use education to build resilience. Use the financial system to disrupt predators at their source.
The goal should not be to remove young people from the digital world. The goal is to make the digital world safer for them to inhabit.
We cannot protect children by isolating them. We protect them by removing the threat.
If You Want More Work Like This
I write The Intelligent Playbook to help people understand how to use AI with confidence, clarity, and purpose. If this topic matters to you, consider subscribing so you can follow future deep dives and practical guides.
Hopefully, this is part of a negotiated process that ends up somewhere productive. Without action these companies will continue to just do whatever gets them the most revenue regardless of the damage they cause.