Australia's under-16 social media ban cut accounts 10 points, but total use barely moved

Three months into Australia's under-16 social media ban, the share of children holding an account on a named restricted platform fell from 52.4% to 42.1%. The share who actually accessed an age-restricted service barely moved, from 85.9% to 81.5%, because most of the platforms on the restricted list do not require an account to view content.

Australia wrote a law that regulates one specific behavior, account creation on a closed list of ten services, and publicly sold it as a different behavior: reducing how much social media minors use. The eSafety Commissioner, Australia's federal online safety regulator, released the first wave of its two-year longitudinal study this month, with the two metrics moving in opposite directions at exactly the moment the policy was supposed to align them.

YouTube sits on the restricted list, and 73.6% of children aged 10 to 15 reported accessing it in the prior four weeks, mostly without an account, by simply watching. Twitch, also restricted, similarly lets viewers in without logging in, because the law was built to restrict account creation, not view access.

TikTok, Snapchat, and Instagram account shares all fell by enough to drive the headline 10-point drop. At least a quarter of children still reported using each of those three services in the prior four weeks, a reminder that account removal and use ending are different things. A third of children in the study reported having an account shut down or deactivated because of the ban, evidence the age-assurance machinery is operating at scale. Reddit held flat at 3.6%, the least-moved data point on the list, while smaller services like Threads (1.5% to 2%) and Kick (0.2% to 0.6%) still grew, suggesting enforcement friction is uneven across the ten platforms.

Nearly 22% of children reported using Pinterest, a platform not on the restricted list, and survey responses point to kids using YouTube and Twitch via search, embedded clips, and shared links rather than accounts. The regulation raised the cost of one entry point and the cohort found cheaper ones.

The eSafety Commissioner also found no statistically significant change in AI chatbot use among the cohort. Chatbot services were not on the restricted list and were not a target of the legislation, so the result is a reminder of the policy's perimeter rather than a measure of its success.

The regulatory category the Australian law created, accounts on a list, and the goal it was publicly sold as serving, reducing minors' social media use, were never the same thing. The instrument worked as written. Other countries copying the Australian template should expect both the headline account drop and the accountless-viewing displacement unless they expand the regulatory surface, for instance by treating view access itself as the regulated behavior or by closing the non-listed-platform gap with a broader designation power.

The eSafety Commissioner is tracking more than 4,100 children and families across the two-year study, with the next data wave due in 2027, after a full year of the law in force and after age-assurance systems have had more time to harden. The metric to watch is whether the access rate, and the displacement to accountless viewing and to non-listed platforms, holds near 81.5% or finally starts to bend.