Skip to content

Original privacy and security research, news and free tools since 2018, cited by the FBI, CISA and the World Economic Forum See our research

Cited by the FBI, CISA and the World Economic Forum

Guide

How to Protect Your Child Online: The Settings That Matter, the Real Risks, and What Apps Collect About Kids

Published Updated

Miklos Zoltan

By Miklos Zoltan

Founder - PrivacyAffairs

Laura Bede

Fact-Checked this

3 Comments

Turn on the free parental controls already built into your child’s phone, tablet or console, then check the teen settings in each app they use.

You don’t need to buy anything for either. Apple, Google, Microsoft and Nintendo include their controls at no cost, and the big apps now put under-18s into restricted teen accounts by default. Instagram, TikTok, Discord and Roblox all tightened theirs in 2025 and 2026.

None of those tools lets you read your child’s messages. They show who your child talks to, not what they say. The rest of the protection is a child who tells you when something feels wrong, and settings make that conversation easier without replacing it.

What children actually run into online

The risks change with age, and so do the apps where they show up.

Under 9. Mostly content and spending: autoplay that drifts from cartoons to something stranger, and in-app purchases. For many children a game with chat is their first social network, which is why Roblox now keeps chat off for its 5–8 age group and Fortnite gives under-13s a chat-free account until a parent agrees.

9 to 12. The first phone. In the UK, smartphone ownership rises from 56% of 10-year-olds to 83% of 11-year-olds, the year most children start secondary school. Group chats arrive, and with them the first messages from people your child has never met.

13 and up. Social media at full volume. Four in ten US teens say they are online almost constantly, about nine in ten use YouTube and 64% have used an AI chatbot (Pew Research Center, December 2025).

The threat that has grown fastest is sextortion. The pattern is well documented. Someone posing as a peer starts a chat, moves it to a private app, gets an intimate image and then demands money or more images. NCMEC, the US clearinghouse for these reports, received 1.4 million online enticement reports in 2025, 158% more than in 2024, and an average of 137 financial sextortion reports a day. The FBI says boys aged 14 to 17 are the most common targets. What to do if it happens is at the end of this page; the short version is don’t pay and don’t delete anything.

The quietest risk is data. Apps collect it from children, target ads with it and sometimes leak it, and some of the apps sold to parents are among the worst offenders.

Start with the controls built into the device

They cost nothing, they cover every app on the device, and they are harder for a child to switch off than anything inside an app.

iPhone and iPad. Set your child up in Family Sharing with a child account, which Apple requires for under-13s and offers up to 18. Screen Time then sets app limits, downtime and content limits. For under-13s, Ask to Buy (your approval for every download) and Ask to Browse (your approval for new websites) are on by default, and Communication Safety, which blurs nude images in Messages and FaceTime, is on for every child under 18. iOS 27, out this autumn, adds Time Allowances by app category and Schedules, and makes Communication Safety block violent images too (Apple, June 2026). The new controls need version 27 on every device in the family.

Android phones and Chromebooks. Google Family Link lets you approve apps from Google Play, set daily limits, bedtime and school time, see the device’s location and lock SafeSearch on. It can’t show messages, searches or Chrome history. Teenagers can be supervised too, with their agreement, and they now need a parent’s approval to end supervision until they turn 18; both of you are notified when it stops (Google). Family Link can’t supervise an iPhone.

Windows and Xbox. Microsoft Family Safety sets screen time, blocks apps, games and websites across Windows, Xbox and Edge, and caps spending.

Nintendo Switch. The free Parental Controls app sets daily play time, game ratings and messaging limits. On Switch 2, children under 16 can use GameChat only with friends you approve.

If you set only one thing, make it app approval: Ask to Buy on Apple, Play approvals in Family Link. It stops the new chat app you’ve never heard of, and it stops the free VPN app installed to get round everything else.

App by app: what teen accounts do and what you can see

Most big apps now put under-18s into a restricted teen account by default. That only works if the app knows the account belongs to a teenager, and most still take the birthday a child types in. A child who claimed to be 18 at age 11 gets the adult version at 13.

Platforms now try to catch that from behaviour. YouTube estimates age from viewing history in the US, UK, EU and several other countries, Instagram uses AI to look for teens on accounts with adult birthdays, Discord estimates an age group from account activity, and Roblox requires a face-based age check before anyone can chat. The simplest fix is older: create the account together, with the real birth date.

Defaults for under-18s, and what each parent tool shows
App
Minimum age
Default for teens
Parent tool
Instagram
13
Private account; messages only from people they follow; 13+ content setting; sleep mode 10 pm–7 am. Under-16s need a parent to loosen any setting.
Supervision: who they messaged in the past 7 days, daily time limits. No message content
TikTok
13 (US under-13s get a separate kids’ version)
Private account; no direct messages under 16; 60-minute daily screen-time limit.
Family Pairing: settings, time limits, who they follow and who follows them. No message content
YouTube
13 (higher for a Google Account in some EU countries)
Break and bedtime reminders on, no personalised ads, age-restricted videos blocked.
Supervised account under 13; supervised teen account that either of you can end, showing uploads, comments and subscriptions.
Snapchat
13
Private account; only friends and saved contacts can message; location on Snap Map off.
Family Center: friends list, who they talked to recently, average daily time. No message content
Roblox
Accounts by age
Roblox Kids (5–8): chat off. Roblox Select (9–15): chat opens gradually, only after a face-based age check.
Parental Controls: friends, time and spending limits until 13; chat and games until 16.
Discord
13
Since Sep 2026, for teens and anyone not age-confirmed: messages from non-friends go to requests, age-restricted servers blocked.
Family Center: new friends, who they messaged or called in the past 7 days, servers joined. No message content
WhatsApp
13 (younger only on a parent-managed account)
Parent-managed accounts for pre-teens, rolling out since Mar 2026: messages and calls only.
Parent PIN: who can contact them and which groups they join. Chats stay encrypted. No message content
Fortnite
Under 13: restricted account
No voice or text chat and no real-money purchases until a parent consents.
Epic Parental Controls: chat limits, PIN for friend requests and purchases, age-rating cap, weekly play limits.
Teen defaults apply when the app knows the user is under 18. Ages can be higher where local law says so.
Sources: Instagram and Meta, TikTok, YouTube and Google, Snap, Roblox, Discord, WhatsApp and Epic Games help and newsroom pages. Checked 18 Nov 2026.

The last column has one thing in common. None of these tools shows message content: Instagram, Snapchat and Discord show who your child talked to, and WhatsApp’s new parent-managed accounts for pre-teens keep chats end-to-end encrypted, so even the parent can’t read them. That’s deliberate. A parent tool that opened every chat would also open every chat to whoever took over the parent’s account.

What apps collect about your child

Children’s data is worth money to advertisers, and the law has only lately caught up. In the US, the amended COPPA Rule has applied since 22 April 2026. Apps and sites aimed at under-13s now need a parent’s separate consent before they share a child’s data with third parties, including for targeted ads; they need a written security program, and they can’t keep children’s data indefinitely (FTC). It still covers only under-13s. COPPA 2.0, which would extend it to teenagers, passed the Senate in March 2026 and is not law.

The biggest cases show what companies did before:

  • TikTok agreed in August 2026 to pay $400 million to settle a US Justice Department children’s privacy case; the claims are allegations, with no finding of liability (DOJ).
  • Epic Games paid a $275 million COPPA penalty in 2022, partly for leaving voice and text chat on by default for children in Fortnite.
  • YouTube paid $170 million in 2019 for tracking viewers of children’s channels to target ads.
  • In Europe, our GDPR fines tracker lists eight fines against apps and platforms over children’s data since 2021, €787 million together. The two biggest were Instagram (€405 million, for showing teenagers’ phone numbers and email addresses on business accounts) and TikTok (€345 million, for setting children’s profiles to public by default).

The apps sold to parents for watching children carry a risk of their own. An app that secretly copies a child’s messages, photos and location to its server is the same product as the spyware sold for watching partners, and those companies leak. TechCrunch counts at least 27 of them hacked or leaking data since 2017, among them mSpy, Cocospy, Spyic and Catwatchful, which was sold as a child-monitoring app. Every message such an app copies sits on a server you don’t control. The controls built into phones and apps copy no messages at all, and they tell your child they’re supervised, which is the fairer deal anyway.

One small setting helps on Apple devices: when an app asks for your child’s age, iOS can share an age range instead of the birth date.

Router filters, and how children get around them

A family DNS filter on your home router blocks adult sites for every device on your Wi-Fi, for free. Replace the DNS servers in your router’s settings with one of these pairs:

  • Cloudflare 1.1.1.1 for Families: 1.1.1.3 and 1.0.0.3, which block malware and adult content (Cloudflare)
  • OpenDNS FamilyShield: 208.67.222.123 and 208.67.220.123
  • CleanBrowsing Family Filter: 185.228.168.168 and 185.228.169.168, which also block VPN and proxy sites

It’s a blunt tool with holes. A phone on mobile data never touches your router. A browser using encrypted DNS (DNS over HTTPS) can skip the router’s DNS, although Firefox switches it off when it detects parental controls. A VPN tunnels past the filter altogether. And what the router can show you is limited; our guide to what parents can see on Wi-Fi covers it.

Children know all this. In a survey for the UK government published in July 2026, 22% of 11- to 17-year-olds said they had used a VPN in the past three months, and 39% had got round at least one age check. The government has said it won’t ban VPNs.

A VPN doesn’t make a child safer, whatever its app store listing says. It hides traffic from your router, not from the apps your child uses, and a free VPN app is one more company handling your child’s browsing. Screen Time has no VPN switch, but a VPN app still needs your approval to install, which is why app approval matters more than any filter. If you find one anyway, ask what it was for before you delete it. The answer usually tells you which site or game your filter blocked, and whether the block made sense.

What the law now requires, country by country

United States. COPPA protects under-13s. No federal law covers older teenagers yet: the Senate passed COPPA 2.0 in March 2026 and the House passed its own package, the KIDS Act, in June, and the two haven’t been reconciled. States move faster, with age checks for adult sites and app stores; our state-by-state tracker lists them.

United Kingdom. Since 25 July 2025, sites with pornography or with self-harm, suicide and eating-disorder content must use highly effective age checks. By September 2026 Ofcom had fined eight porn site operators and 4chan for missing them. On 15 June 2026 the government announced a social media ban for under-16s covering Snapchat, TikTok, YouTube, Instagram, Facebook and X but not WhatsApp or Signal, expected from spring 2027.

European Union. The Commission’s guidelines under the Digital Services Act (July 2025) tell platforms how to protect minors, and member states are meant to offer the EU age-verification app by the end of 2026. On 17 September 2026 the Commission proposed an EU-wide rule: no social media under 13, and accounts of their own only from 15. It is still a proposal. France’s under-15 ban never took effect; on 14 August 2026 the Constitutional Council struck it down as disproportionate, partly because every adult would have had to prove their age.

Australia. Since 10 December 2025, Facebook, Instagram, Snapchat, TikTok, YouTube, X, Threads, Reddit and Kick must keep under-16s off. By mid-January 2026, 4.7 million accounts had been deactivated, removed or restricted. Malaysia’s under-16 rule followed on 1 June 2026.

Here I’ll take a side. Most of these laws protect children by making everyone prove their age, which means millions of adults handing IDs and face scans to apps and verification vendors, the kind of companies whose data leaks fill our news section. The French court saw the problem. There is a better design already in our phones: the parent sets the child’s age once on the device, and apps learn only “under 16”, never a name or a passport number. Lawmakers should require that before they require one more upload of your driver’s licence. A national database of IDs built to protect children sounds safe right up to its first breach. How online age checks work explains the methods in use today.

If something goes wrong

If your child is being blackmailed over an image. Don’t pay; NCMEC says paying or cooperating rarely stops it. Stop replying and block the account, but don’t delete the account or the messages, and screenshot the username and the demands. Don’t copy or forward the image, and don’t ask to see it: UK guidance says copying or sharing it is illegal, and the Internet Watch Foundation advises parents not to ask.

Then report it:

  • US: NCMEC’s CyberTipline at report.cybertip.org or 1-800-843-5678.
  • UK: CEOP, the National Crime Agency’s child protection unit (call 999 if your child is in danger), and Report Remove from the IWF and Childline, which young people can use themselves.
  • Elsewhere: the national hotline in your country from the INHOPE network, or the police.

Take It Down helps stop the image spreading. It works for images taken before 18: it makes a digital fingerprint of the image on your device, so the image itself is never uploaded, and participating platforms (Instagram, Facebook, TikTok, Snapchat and YouTube among them) use it to block copies. It can’t reach encrypted chats. The same routes apply to fake nude images made with AI; our deepfakes guide covers removing those.

If a stranger has been contacting your child. Keep the messages, report the account in the app and block it. If there’s any talk of meeting, or anything sexual, report it to the police or the hotlines above as well.

If your child is being bullied. Screenshot, report in the app, block, and tell the school if classmates are involved.

All of this is easier if your child comes to you early. A child who expects to lose their phone for admitting a mistake has every reason to stay quiet, so say it in advance: telling you about something that went wrong online won’t get them punished.

Frequently Asked Questions

Some people found answers to these questions helpful

What age should a child get a smartphone?

No law sets one. Most social apps require users to be 13, and Apple requires a child account under 13, so a younger child's phone can run with Screen Time or Family Link and your approval for every app. In the UK most children get one at 10 or 11, around the move to secondary school. What's on the phone matters more than the birthday.


Can a 10-year-old have a YouTube account?

Not their own. YouTube's terms require users to be 13, and some European countries set a higher minimum age for a Google Account: 14 in Spain, Italy and Austria, 15 in France, and 16 in Germany, the Netherlands and Ireland. Below that age, a parent can set up a supervised YouTube account or YouTube Kids. A supervised child can watch but can't comment, upload videos or use live chat.


Should I read my child's messages?

The parent tools won't let you: Instagram, Snapchat and Discord show who your child talks to, not what they say. Apps that promise full access are spyware with a parental label, and many have leaked what they collected. With younger children, approving friend requests and keeping phones out of bedrooms at night does more. With teenagers, reading messages in secret tends to end the conversations you most need them to have with you.


Can kids get around Screen Time or Family Link?

Some do, with a VPN, a second device, a friend's phone or mobile data that skips your home filter. In the UK survey cited above, 39% of 11- to 17-year-olds had got round at least one age check. Switching supervision off is harder: teenagers on Family Link need a parent's approval to stop it until 18, and on Apple a 13- to 17-year-old can't leave Family Sharing without the organiser's permission once Screen Time is set up.


Does a VPN make my child safer online?

No. A VPN hides traffic from your internet provider and your router, which is exactly why children use one to get past family filters. It does nothing about who contacts them in an app or what the app collects. If you find a VPN app on your child's phone that you didn't install, treat it as a conversation to have rather than a setting to keep.


What should I do if my child is being blackmailed with a photo?

Don't pay, stop replying, block the account but keep the messages, and report it: in the US to NCMEC's CyberTipline, in the UK to CEOP, elsewhere to the police. Use Take It Down to stop the image spreading on the major platforms. Tell your child it isn't their fault; NCMEC received more than 50,000 financial sextortion reports in 2025, so they are far from the only one.


3 Comments

  • MOHAN

    April 28, 2019 3:47 am

    With opera vpn or the add on of avira phantom vpn is also a nice choice for protecting privacy protection.

  • Klara

    March 25, 2019 4:32 pm

    Thank you for this detailed guide. I will be sharing this to my friends as well. It is very well written and it’s obvious the author is a mother herself.

    • VPNTeacher

      March 25, 2019 4:38 pm

      Thank you. Elizabeth indeed worked really hard on this and is writing from experience 🙂