On 10 December 2025, Australia became the first country in the world to legally bar children under 16 from holding social media accounts. The Online Safety Amendment (Social Media Minimum Age) Act 2024 — passed by the Albanese government in November 2024 after a single day of substantive Senate debate — places the compliance burden squarely on platforms. Meta's Instagram, Facebook and Threads, along with TikTok, Snapchat, X, Reddit and YouTube, must now take 'reasonable steps' to detect and deactivate underage accounts or face civil penalties of up to A$49.5 million per breach, enforced by eSafety Commissioner Julie Inman Grant.
Nearly six months in, the Australian experiment has become the most-watched policy laboratory in online safety. The United Kingdom, France, several US states, and Brussels are all studying the rollout closely. The early lessons are instructive — and not flattering.
A Law Built on a Technological Wager
The Act does not prescribe how platforms must verify age. That deliberate ambiguity was meant to encourage innovation; in practice it has shifted a profound design question — how do you reliably tell a 15-year-old from a 16-year-old at internet scale? — onto private companies operating under the threat of nine-figure fines.
The government-commissioned Age Assurance Technology Trial, run by the UK-based Age Check Certification Scheme and delivered to the Department of Communications in mid-2025, found that no single method is both accurate and privacy-preserving across the population. Facial age estimation tools were shown to have meaningful error margins, particularly for users aged 13 to 17 and for users with darker skin tones. ID-document checks work but exclude the roughly one in five Australian teenagers who lack a government-issued photo ID. Behavioural inference — guessing age from typing patterns, contacts and content — is the most invisible and arguably the most invasive.
Platforms have rolled out a patchwork: Meta is using Yoti's face-scan estimation alongside ID uploads; TikTok and Snapchat are leaning on third-party providers and parental verification flows; YouTube is restricting account creation while leaving signed-out viewing largely intact. Reports from Australian users through early 2026 describe waves of account deactivations, appeals backlogs, and a rising secondary market in VPNs and borrowed adult identities.
The Civil-Liberties Backlash
Opposition has been broad and unusually cross-partisan. Digital Rights Watch has called the law 'a privacy disaster wrapped in a child-safety bow,' arguing that universal age assurance effectively means universal identity verification — every adult Australian must now prove they are an adult to use mainstream platforms. The Australian Human Rights Commission warned during parliamentary hearings that the bill risked breaching Australia's obligations under the UN Convention on the Rights of the Child, particularly Article 13 on freedom of expression and Article 17 on access to information.
The exclusion problem is real. For LGBTQ+ teenagers, young people in remote communities, and children in unsafe homes, social platforms are often the primary route to support networks, identity exploration and emergency information. Cutting off that channel through age-gating, without robust offline alternatives, is not a neutral act.
The Innovation Cost
From a pro-innovation perspective, the more troubling pattern is structural. The Act's compliance demands are easily absorbed by Meta, ByteDance and Alphabet — companies that can afford to build, license and litigate around age-assurance systems. They are crushing for the smaller forums, hobbyist communities and Australian-grown platforms that the same government claims to want to nurture. Reddit, notably, has signalled it may simply restrict Australian access to certain communities rather than try to engineer a compliant teen experience.
This is the recurring lesson of the past decade of platform regulation, from Germany's NetzDG to the EU's Digital Services Act: prescriptive rules written with a handful of US giants in mind end up entrenching them. Australia's Act now adds a global age-verification layer to that pattern.
A Better Way Forward
None of this is to deny the underlying problem. The mental-health evidence on heavy adolescent social media use, while contested, is concerning enough to justify serious policy responses. But the proportionate options were never properly tried.
- Default-on safety settings for minors — private accounts, restricted DMs, no algorithmic amplification — which the eSafety Commissioner already had powers to mandate under the Basic Online Safety Expectations.
- Device-level age signals shared with platforms via the operating system, an approach Apple and Google have both proposed and which keeps biometric data off third-party servers.
- Stronger enforcement of existing rules against design features known to harm minors — autoplay, infinite scroll, streaks, engagement-bait notifications.
Each of these would have advanced child safety without requiring every adult Australian to hand over a face scan or driver's licence to read a tweet.
What Comes Next
The eSafety Commissioner has signalled that the first formal compliance assessments will come in the second half of 2026, and the Act includes a statutory review after two years of operation. By then, the evidence base will be considerably richer — on circumvention rates, mental-health outcomes, the cost of false positives for young adults wrongly locked out, and the privacy footprint of mass age assurance.
Other governments should resist the temptation to copy-paste the headline before the data is in. Banning teenagers from the open internet is a serious thing to do. It deserves more than a year of evidence before becoming the global default.