Australia's Online Safety Amendment (Social Media Minimum Age) Act 2024 was the first law of its kind anywhere: platforms must take "reasonable steps" to keep under-16s off their services or face penalties. It received royal assent on 10 December 2024 and took effect exactly a year later, on 10 December 2025. Eight months on, the regulator tasked with enforcing it has published the first hard data on whether it is working — and the results complicate the case for the model Australia is now exporting.
The Case For The Ban, Fairly Stated
The strongest argument for a minimum age wasn't that it would eliminate teen social media use overnight — its architects always framed this as a multi-year behavioural shift, not a light switch. Removing algorithmic feeds and open DMs from the youngest, most vulnerable users for even a few extra years reduces cumulative exposure to harm during a developmentally sensitive window, even if determined teenagers eventually find workarounds. That's a real public-health logic, not a strawman, and it's why eSafety designed a two-year longitudinal evaluation — run with Stanford University's Social Media Lab and an 11-member international academic advisory group — rather than declaring victory or failure after one quarter.
What The First Report Actually Found
That evaluation's first instalment, published 31 July 2026, surveyed 803 children aged 10–15 and their parents, comparing a pre-ban baseline from December 2025 against a March–April 2026 follow-up. The headline number: use of at least one age-restricted platform fell only marginally, from 85.9% to 81.5%. Account ownership among under-16s dropped more substantially, from roughly 52% to 42% — evidence platforms did remove some accounts. But owning fewer accounts and using platforms less are different things, and the gap between those two figures is the story: a large share of children without a registered account are still accessing the same services, logged in as someone else, on a sibling's device, or simply never challenged at all.
That last point is the law's central weakness. Over half of the surveyed children said platforms never asked them to verify their age in the first place. A statute built entirely on the premise that companies will take "reasonable steps" to check age is only as strong as the checking — and eSafety's own data shows the checking mostly isn't happening.
The Parental-Awareness Paradox
The more striking finding, and the one regulators advocating similar bans elsewhere should sit with, is that parental awareness of their children's social media use fell rather than rose. The evaluation found the share of parents who didn't know their child had used social media in the reporting period increased by roughly 10 percentage points after the ban took effect. A policy sold partly on giving parents more visibility and control appears, three months in, to have done the opposite — plausibly because children who previously used social media openly, with parents managing settings and time limits, are now doing so quietly to avoid the appearance of breaking the rules. Prohibition without effective enforcement doesn't just fail to reduce the underlying behaviour; it can push it further from adult supervision, which is a worse outcome by the ban's own stated goals.
Migration, Not Elimination
The evaluation also found early signs of displacement toward platforms outside the restricted list, including Reddit and Pinterest, and "little to no change" in children's offline activity — undercutting the theory that time freed from Instagram or TikTok would translate into more outdoor play or family time. As other regulators have separately noted, facial and behavioural age-estimation tools remain unreliable near the 16-year threshold in either direction, meaning the compliance burden platforms face is real even where enforcement is weak, while the protective benefit to children remains, so far, unproven.
Why This Matters Beyond Australia
The UK, Indonesia, Malaysia and several EU member states are watching Canberra's experience as they weigh their own age-gating proposals, and the EU Commission has signalled plans for a similar model later this year. Australia's evaluation is exactly the right instrument for that debate — transparent, longitudinal, methodologically serious — and eSafety deserves credit for publishing an inconvenient first result rather than waiting for friendlier numbers. But the early data argues for caution, not imitation: a law can look decisive in a press release and still fail on its own metrics when the enforcement mechanism — age verification — is the weakest link in the chain. Lawmakers elsewhere should treat this report as a warning about implementation, not a referendum on whether youth online safety is worth pursuing. The honest reading of Australia's first quarter is that banning accounts is easy to legislate and hard to make real; enforcement design, not the ban itself, is the unsolved problem.