A Genuine Problem, a Blunt Instrument
Six months into Australia's world-first social media minimum age regime, the numbers are uncomfortable for its architects. A peer-reviewed study published in The BMJ in June 2026, drawing on surveys of more than 400 young Australians before and three months after the December 10, 2025 enforcement date, found that over 85 percent of under-16 participants were still accessing banned platforms. Seven in ten of those who remained online said it was "easy" to circumvent the restrictions.
The case for acting is genuine and should not be dismissed. There is a growing body of evidence — imperfect but real — linking unstructured adolescent social media use to anxiety, sleep disruption, and exposure to self-harm content. Australia's Online Safety Amendment (Social Media Minimum Age) Act 2024, which passed Parliament on November 29, 2024 and entered enforcement in December 2025, reflects a legitimate state interest in drawing a line. The government's frustration with platforms that have absorbed years of regulatory warnings without meaningful design changes is equally understandable.
But on June 27, 2026, Prime Minister Anthony Albanese responded to the evidence of non-compliance by doubling the maximum platform fine from A$49.5 million to A$99 million and expanding the eSafety Commissioner's powers to compel platforms, age-verification providers, and app stores to hand over compliance evidence. The diagnosis the government is offering — big tech isn't trying hard enough — may be incomplete. The deeper problem is that the law's "reasonable steps" standard left the architecture of enforcement undefined, and higher penalties for an undefined obligation don't automatically produce better outcomes.
What the Compliance Data Actually Shows
The government's own figures are not uniformly bad. Since December 2025, platforms have removed, deactivated, or restricted access to more than five million under-16 accounts across Australia. Five platforms — Facebook, Instagram, Snapchat, TikTok, and YouTube — are under active eSafety investigation. Communications Minister Anika Wells characterised the industry's response as "tricks straight out of the big tech playbook" and "doing the bare minimum to get by."
The BMJ study, however, suggests the problem runs deeper than platform bad faith. The most common age-verification methods encountered by under-16 users were self-declared age (24–39 percent of encounters) and photo uploads (13–27 percent). Both are trivially defeated. Fake accounts accounted for 15–19 percent of circumvention; private browsing for 6–11 percent. Among 12–13 year-olds, daily social media use was essentially unchanged three months after the ban. Among 14–15 year-olds, there was a modest decline in daily use — from roughly 78 percent to 69 percent — which could signal early policy effect, though the researchers caution the sample was drawn from a single state, New South Wales, and relied on self-reporting.
The BMJ authors' precise framing matters: they found "insufficient evidence to conclude that exposure to the Act had any substantial effects." That is different from evidence that the Act has failed. Early-period data from a rolling compliance regime, with platforms still building and deploying age-assurance tools, is genuinely hard to interpret. But the circumvention rate — easy-to-bypass photo uploads and self-declaration dominating the field — is not a signal that platforms need more financial pressure. It is a signal that the technical baseline was never set.
The Fine Is Not the Architecture
Here is the core policy mismatch: A$99 million sounds large, but for platforms whose Australian revenue is a rounding error in global accounts, the marginal deterrence of moving from A$49.5M to A$99M is limited. TikTok's global revenue in 2025 exceeded US$30 billion. Meta's Australian ad revenue dwarfs the fine. What moves these companies is not the penalty level — it is the clarity and specificity of the obligation.
The Act's "reasonable steps" standard was intentionally flexible, avoiding mandated technology so as not to prescribe surveillance solutions. That instinct was correct. Mandating biometrics or government-ID checks would create its own civil liberties crisis, and the Office of the Australian Information Commissioner's co-regulatory role — enforcing Section 63F of the Online Safety Act 2021, which requires platforms to destroy age-assurance data and restrict its use to age determination purposes only — wisely addresses that risk.
But flexibility without a technical floor means platforms can point to any age-verification gesture as compliance, even if that gesture is a self-declaration checkbox. Expanding the eSafety Commissioner's information-gathering powers helps — compelled disclosure of compliance evidence is a meaningful tool — but it addresses the evidentiary gap, not the verification gap. What proportionate, effective regulation would look like is becoming clearer from evidence: mandatory minimum technical standards for age assurance (akin to the UK's Children's Code approach under the ICO), design obligations that default to age-appropriate settings, and interoperable infrastructure that allows age verification without requiring platforms to collect new identity data.
The Global Precedent Problem
Australia went first deliberately, and the UK is watching closely, with plans for a similar age restriction operative by spring 2027. Policymakers on both sides of the Atlantic are drawing lessons from the Australian rollout. The lesson Canberra appears to be sending — that the correct response to early evidence of limited effectiveness is to escalate penalties — risks becoming the dominant regulatory template.
A better lesson is available: the verification architecture was underspecified from the start. Fines should rise in proportion to evidence that platforms are acting in bad faith, not in proportion to the difficulty of a technical problem that was never adequately defined. Doubling the penalty before defining the obligation is a political signal dressed as a policy solution. Australia's children deserve a regime that actually reduces their exposure to harm — not one that expresses that ambition more expensively.