Skip to content

Experts insights

Banning Kids from Social Media: Easier Said Than Done

TL;DR

  • The UAE has banned social media for under-15s, joining Australia, the UK, Indonesia, Türkiye, and a dozen more countries moving in the same direction.
  • The clinical evidence behind these bans is serious: adolescent mental health, sleep disruption, grooming, addictive design.
  • Definitions of "social media" do not line up across jurisdictions: WhatsApp, Roblox, Discord, and Telegram sit on different sides of the line depending on the country.
  • Enforcement is failing on two sides at once: platforms cannot reliably verify identity at scale, and adults cannot realistically supervise a generation more digitally fluent than they are.
  • The real answer is not the ban itself but what gets built underneath it: education for children and adults, layered technical controls at network and device level, and possibly a dedicated category of devices for minors.
  • A ban is a starting condition, not a solution.

The United Arab Emirates has joined a growing list of countries deciding that minors should not be on social media. The country decided to ban children under 15 from creating accounts on TikTok, Instagram, Snapchat, Facebook, and X, and gave platforms twelve months to comply. This is the latest move in a trend that started with Australia in 2025 and is now spreading across Europe, the Middle East, and parts of Asia.

The intent is understandable and the evidence behind it is serious, but the real question is whether the ban is enforceable.

Why governments are acting

The motivations are remarkably consistent across jurisdictions, and rest on a decade of accumulating clinical and behavioural evidence:

  • Mental health deterioration in adolescents: rising anxiety, depression, and self-harm, particularly among girls.
  • Sleep disruption, with measurable effects on school performance and physical development.
  • Exposure to predators, grooming, and the normalisation of explicit content among minors.
  • Cyberbullying and the amplification of misogyny, racism, and extremist material through algorithmic recommendation.
  • Eating disorders driven by curated body imagery and pro-anorexia communities.
  • Addictive engagement patterns engineered into the platforms themselves, drawing on the same behavioural principles used in gambling.
  • Erosion of attention, real-world relationships, and the developmental experiences that childhood is supposed to include.

Adhering countries

The wave is broad, and the picture is best understood by combining age threshold and legislative status:

Penalties vary in form but converge on severity. Australia has set fines up to AUD 49.5 million for non-compliant platforms, and roughly five million accounts were deactivated in the first three months. The UAE assigns oversight to the National Media Authority and the Telecommunications and Digital Government Regulatory Authority, both empowered to issue warnings, impose administrative penalties, or block platforms partially or fully. Most other jurisdictions follow a similar logic: administrative sanctions first, escalating to access restrictions, with the threat of outright blocking as the ultimate lever.

On paper, this is a serious regulatory framework, but, in practice, the gap between regulation and enforcement is where the trouble starts.

What counts as “social media”?

The current frameworks are not clearly aligned when listing the covered applications, adding more confusion to the matter.

Australia's law covers Facebook, Instagram, Snapchat, Threads, TikTok, X, YouTube, Reddit, Twitch, and Kick, excluding WhatsApp and YouTube Kids. The UK proposal excludes WhatsApp as well, and Signal. Indonesia, on the other hand, includes Roblox, which most parents would call a game rather than a social media. The UAE resolution names only five applications: TikTok, Instagram, Snapchat, Facebook, and X.

The boundary of what counts as "social media" is genuinely blurred:

  • Messaging platforms (WhatsApp, Telegram, Signal, iMessage) carry group chats of hundreds of users, support broadcast channels, and host content sharing that is functionally indistinguishable from a social feed. Telegram in particular operates large public channels that look much more like X than like a private chat.
  • Online games with chat features (Fortnite, Roblox, Minecraft) expose children to communication with strangers, algorithmic content, and behavioural design patterns very similar to social networks.
  • Discord sits somewhere between messaging, gaming, and social media, hosting public communities of millions of users alongside private group chats, but rarely considered in the social media ban.
  • Live-streaming and short-video platforms (Twitch, Kick) are very appealing to kids but barely included in the frameworks.

Until governments commit to a clear, technically grounded definition of what they are banning, platforms will spend the transition period arguing about which side of the line they sit on, and children will simply migrate to whatever falls outside it.

The enforcement problem

Age verification at the scale of a billion users is genuinely hard. The methods approved by most frameworks (government digital identity, ID document scanning with a face match, biometric matching, AI-based age estimation) are well known but also not hard to defeat:

  • A child can use a parent's, older sibling's, or friend's ID: the face match confirms only that a face matches a document, not that the document belongs to the person presenting it.
  • AI-based age estimation has known error rates particularly around 13 to 16. A teenager who looks older, or who learns which photos pass the check, can get through.
  • VPNs can route traffic through jurisdictions where the ban does not apply, and most teenagers know how to install one.
  • Smaller platforms outside the regulatory perimeter can absorb displaced traffic.
  • Mass identity verification also creates its own risk surface: face scans, ID documents, and biometric data become valuable targets for attackers, and a new category of sensitive data ends up in vendor databases that were not built to hold it safely.

The enforcement problem is also proven by early evidence from Australia: according to the country's e-Safety Commissioner, seven out of ten children under 16 who held an account before the ban still have some form of access.

Supervision is hard

Even if platforms managed to keep underage users off their services, there is a second enforcement layer no regulator has properly addressed: the adults supposed to supervise.

A ban implicitly assumes that parents, teachers, and educators can verify that children are actually staying away from the prohibited apps. That assumption falls apart on contact with reality.

Children today receive their first smartphone often before they are ten, and by the time they reach the threshold of any ban, they have years of technical fluency most adults around them do not. Apps can be hidden, renamed, disguised as utilities, or installed and uninstalled in seconds.

Parental control software, on the other hand, is fragmented across operating systems, requires configuration most parents never complete, and is routinely bypassed.

Schools face the same problem at scale. A classroom of thirty teenagers with phones in their pockets cannot realistically be monitored for app-level compliance without application-layer firewalls that block specific apps at the network level. Even solving that problem, the challenge shifts to distinguishing between adult and kids accounts on the same network, or to implement different networks for teachers and students. The same applies to home computers and tablets, where the supervising adult is not a technician and shares the network with the child.

Which solutions?

If neither platform-level verification nor adult supervision can carry the full weight of enforcement, then the conversation needs to move to the practical measures.

Three directions deserve attention, and they work best in combination:

  • Education, for children and for adults. Children need to understand why the rules exist before they will respect them: how recommendation algorithms work, why infinite scrolling exists, what behavioural design is doing to their attention, and what the consequences of breaking the rules are. This presupposes that the adults around them are themselves informed. A rule enforced by adults who do not understand what they are enforcing will not survive contact with a determined teenager.
  • Technological controls at network and device level. The technical toolkit already exists but is not deployed coherently: application-layer firewalls on home and school networks blocking traffic to prohibited platforms; parental control apps configured to restrict installation, browsing, and account creation; whitelisting models where new apps require explicit adult approval rather than blacklisting models that always lag behind; DNS-level filtering that catches workarounds bypassing on-device controls. None is a silver bullet, but combined, they raise the technical bar high enough that circumvention requires real effort rather than a thirty-second tutorial.
  • Dedicated devices for minors. If verifying identity at the platform is unreliable and supervising usage at the device is impractical, the enforcement layer that remains is the device itself. A dedicated category of devices for under-16 use, with social media blocked at the operating system level, sideloading disabled, and recognisable physical or visual markers that allow any adult to identify the device at a glance. Coverage would need to extend across smartphones, tablets, and laptops, both personal and school-issued, with a clean upgrade path at the age threshold. It requires coordination between hardware manufacturers, operating system vendors, schools, and parents that does not currently exist, and it raises real questions about cost, equity, and stigma. None of these problems is insurmountable, but none is being seriously discussed either.

None of these layers works on its own, but combined, they define the actual principle: a ban is a starting condition, not a solution. The real answer lies in the education, technical, and device layers built underneath it, and anchored in a mature sense of technological accountability.

Clearer decisions. Better budgeting. Fewer losses.

From "how many attacks?" to "how much damage?"
The clarity needed for sharper strategy.