_An update to the 2023 post on children and social media. The third of five 2026 updates to the children-focused posts in the privacy series._

When I wrote about children and social media in June 2023, the conversation parents were having was if and when to allow access to these platforms. The framing was about negotiating with an industry whose products were designed against the parent and the child both. Three years later, almost every variable in that conversation has shifted. The platforms have changed. The regulators have moved. The policy environment has produced concrete proposals nobody was discussing in 2023. The grassroots position has changed. And the underlying public health argument has been settled, in a way that closes off several positions parents used to take.

This post is the 2026 version of the conversation. The 2023 framing — staged ages, platform-by-platform settings, parent as advisor rather than surveillant — is still the right framing. The defaults inside it have moved.

What is different in 2026

Five shifts of substantial weight since 2023.

The Online Safety Act took effect. The OSA's age-assurance provisions began to bite during 2024 and 2025, starting with pornographic content services and extending into social platforms. Ofcom is now the active regulator of online safety. The platforms have responded; the responses are uneven and the enforcement is gradual. The bar for what platforms must do to identify minors and provide age-appropriate experiences has risen.

Australia banned social media for under-16s. The Australian government legislated a ban on under-16 social media accounts in December 2024, with enforcement starting late 2025. The data on whether it has worked is contested. The Norwegian government followed in early 2025 with a more limited version. The UK has not legislated a ban and is unlikely to in this Parliament. The conversation about whether such a ban is right has, in two years, moved from fringe to mainstream policy debate.

Smartphone Free Childhood and the delay-first-phone movement. Daisy Greenwell and Clare Reynolds founded Smartphone Free Childhood in early 2024. The movement scaled faster than almost any other parent-advocacy movement in living memory. By 2026 most UK primary schools have associated parent groups committing to delay first smartphones to age 14, often year-group-wide so the social pressure on the early adopters is reduced. The change is visible at the school gate.

Phones out of school. The DfE's February 2024 guidance has been adopted at most UK secondaries. The school day, in 2026, is increasingly phone-free.

The public-health argument has been settled. The Jonathan Haidt argument — that smartphones and social media have caused a measurable adolescent mental-health crisis — was contested in 2023. By 2026 the contested position has narrowed; the broad consensus has shifted toward yes, the platforms have done significant harm, particularly to adolescent girls, and the mechanisms are well-documented. The remaining debate is what to do about it, not whether the harm is real.

The platforms in 2026

A quick map.

TikTok remains the dominant short-video platform among British secondary children. The ownership question that hung over it through 2023-25 has, after various twists, resolved into a structure under which the US operations are partly separated. The UK operation continues. Ofcom enforcement against the platform has resulted in materially stronger age-defaults for under-18s. Teen accounts default to private, restrictive content filtering, and screen-time limits. The platform is not what it was in 2023 for under-16s.

Instagram and Meta — Instagram introduced Teen Accounts in late 2024 — private by default, restricted DMs, parent-supervised messaging. The reality is that the controls work better than the previous generation; they are not perfect. The Threads platform Meta launched in 2023 is now a Twitter-substitute with similar dynamics, less commonly used by children.

Snapchat remains the dominant peer-messaging platform. The platform has been the subject of multiple coroner inquests and lawsuits over the past three years over its role in adolescent harm. Product changes have followed; the underlying ephemeral-messaging model that creates much of the risk is unchanged.

WhatsApp remains the dominant family-and-friend group messaging platform. The minimum age went up to 13 (from 16 in the UK pre-2024); the practical reality is similar to 2023.

YouTube has changed least and matters most. The platform that is least thought of as social media is, by viewing time, the dominant influence in many children's lives. The recommendation algorithm remains a significant concern. YouTube Kids has matured for younger children; the main platform's age-appropriate experiences for older children are improved but still patchy.

Discord has rolled out age-assurance and parental controls through 2024-25. The platform is now significantly safer at the platform level than it was. The risk has shifted to specific servers and the cultures within them, which the platform cannot police centrally.

Bluesky emerged as a Twitter-alternative through 2023-24 and has matured into a smaller but distinct social platform. Decentralised and ActivityPub-style federation across Mastodon and similar. Smaller children's-safety footprint but the federation model has its own implications.

BeReal collapsed and pivoted. Once the platform a year-group might all be on, by 2026 mostly absent.

The staged approach, updated

The four-stage framework from 2023 — pre-teen no social, 11-13 first messaging platform, 13-16 first personal accounts with privacy settings, 15+ increasing autonomy — still works. Two updates.

Stage one (pre-teen) is now longer in many households. The 2023 default was that primary-age children would have access to family-supervised platforms but no personal accounts. The 2026 default among the most engaged parents is to extend that further — no personal smartphone, no personal social accounts, until 13 or 14. The Smartphone Free Childhood position. This is not for every family. It is now available as a position that does not require defending.

Stage three (first personal accounts) has better defaults. The Instagram Teen Account, the TikTok teen account, the Snapchat Family Center, the Discord parental controls. The fifteen-minute setup of privacy settings is now closer to confirm the defaults are sensible than reverse the platform's intentional anti-privacy defaults.

What is still hard

Three things the 2023 post described that are still hard in 2026.

The conversation about intimate image sharing. The platforms have improved their content controls. The peer-to-peer image-sharing dynamics among teenagers have not improved meaningfully. The Childline Report Remove and NCMEC Take It Down tools are mature and used widely. The conversation parents need to have with teenagers is unchanged. Some of it is now AI-shaped — covered in the AI/deepfake update later in this mini-series.

The recommendation-algorithm-as-rabbit-hole problem. The Ofcom code-of-practice work has put pressure on the worst patterns but the algorithmic feeds remain optimised for engagement. The pattern of bright child drifts toward harmful subculture via algorithmic recommendation has not been solved.

The friend-pressure problem. The first child in a year group to have a smartphone applies pressure to the next. The whole-year-group commitment that Smartphone Free Childhood encourages is the answer; it requires social organisation among parents. Where it works, it works well.

What this month looks like

Two short pieces of work.

One: if you have not yet, look up whether your child's primary school year group has a Smartphone Free Childhood pact, and consider joining or starting one.

Two: for any child already on a platform, redo the privacy settings against the 2026 defaults. The platforms have changed enough that the 2023 setup is, in places, no longer current.

Next month: gaming, revisited. The Roblox of 2026 is not quite the Roblox of 2023.