Australia's Under-16 Ban Is Misfiring, and the World Should Not Copy the Blueprint

The design was elegant on paper. When Australia's under-16 social media ban took effect on December 10, 2025, the world's first national prohibition of its kind, the enforcement engine had two moving parts. Platforms would purge existing accounts held by minors, then deploy age-assurance systems to keep new ones from forming. Regulators would audit compliance and levy fines steep enough to make evasion irrational. Seven months later, both parts are slipping their gears, and on July 3, 2026, the Senate pulled the wrench out of the government's hands entirely, referring the bill meant to tighten the machinery to an eight-week inquiry (MLex, US News). The failure is now a live argument in editorial pages from Washington to Sydney over a single question: is this a model to export, or a warning to heed?

Mechanics of a policy that stopped turning over

Understanding why the ban is faltering requires looking at how it was supposed to work. The law does not police children directly. It places the legal burden on the platforms, requiring them to take "reasonable steps" to exclude anyone under 16. That phrasing is the load-bearing beam, and it is proving flexible enough to bend.

In the weeks after implementation, companies reported deactivating more than five million accounts, a number that suggested the first part of the engine was firing. Yet the follow-through told a different story. According to eSafety Commissioner data reported by Fortune on June 29, 2026, and by Al Jazeera on June 27, roughly seven in ten children who held accounts on restricted platforms when the ban began remain on them. Deletion, it turns out, is not the same as exclusion. An account removed in December can be recreated in January by a determined teenager, and the age-assurance checks meant to stop that recreation have been uneven at best.

Split verdict inside the regulator's caseload

Not every platform sits in the same position, which is instructive. The eSafety Commissioner has expressed satisfaction with the steps taken by X, Kick, Reddit, Threads, and Twitch, according to reporting around the June data release. The regulator's frustration is concentrated on five larger services, with investigations underway into Facebook, Instagram, Snapchat, TikTok, and YouTube (Fortune, Al Jazeera). The pattern matters because it undercuts the simplest explanations. This is not a case of technology being impossible or every operator resisting equally. It is a case of the largest incumbents, the ones with the most users and the most to lose, calibrating their compliance to the floor.

Communications Minister Anika Wells put the accusation plainly, saying the major platforms are "adopting tricks straight out of the Big Tech playbook and doing the bare minimum to get by" (Fortune, Al Jazeera). The complaint is candid, and it doubles as a diagnosis of the design flaw. A standard defined as "reasonable steps" invites litigation over what counts as reasonable, and litigation favors the party with the deepest bench of lawyers.

Legislative repair that stalled in the chamber

The government's answer was to harden the incentives. The amendments now sitting in limbo would double the maximum fine for systemic breaches to A$99 million, about US$68 million, and grant the eSafety Commissioner authority to compel information from platforms, app stores, and age-assurance providers, according to Al Jazeera. That second power is arguably the more consequential of the two. Without it, the regulator is asking companies to grade their own homework; with it, the regulator can inspect the machinery directly.

The repair did not reach a vote. Because the governing Labor Party lacks a Senate majority, opposition and crossbench senators were able to refer the bill to a committee for eight weeks (MLex, US News). Prime Minister Anthony Albanese condemned the delay, warning that the interval hands platforms room to maneuver before tougher rules bind. Whatever the merits of that specific charge, the procedural reality is unambiguous. The fix for a policy criticized as toothless has itself been paused, and the gap between announcement and enforcement has widened rather than closed.

Export case that collapses under inspection

Australia was meant to be the proof of concept. Governments in Europe and elsewhere have watched Canberra closely, treating the ban as a template that could be lifted and localized. A Washington Post opinion piece published July 8, 2026, argued directly against that instinct, contending that the failing Australian experiment should not be replicated by other countries.

The argument is worth taking seriously on the merits rather than the mood. The case for caution rests on the same mechanics that are now visibly straining:

• A prohibition that depends on platform-side age assurance inherits every weakness of age assurance, including circumvention through recreated accounts, borrowed logins, and inconsistent verification.
• A "reasonable steps" standard converts a bright-line ban into an open-ended negotiation, and the largest firms are best equipped to win that negotiation slowly.
• Enforcement teeth arrive last and most fragilely, hostage to whatever parliamentary arithmetic happens to prevail when the enabling legislation reaches the floor.

A country importing the model imports all three vulnerabilities at once, without the seven months of field data that Australia has now paid for. That is the substance of the case against copying: not that protecting children online is misguided, but that this particular apparatus has not yet demonstrated it can do the protecting.

Compliance floors versus compliance ceilings

The deeper lesson sits in the phrase Wells reached for. When a regulation sets a behavioral floor and leaves the ceiling undefined, rational firms optimize toward the floor. The Australian design assumed platforms would treat the ban as a mandate to solve; several of the largest instead treated it as a compliance minimum to satisfy. Any government drafting a similar law has to decide, in advance, whether its statutory language rewards genuine exclusion or merely documented effort. Australia is discovering the difference in real time.

Reading the stumble without overreading it

Prudence cuts against triumphalism in either direction. The ban is seven months old, an adolescence for any regulatory regime, and the eSafety Commissioner's investigations into the five lagging platforms have not concluded. It is possible that the compelled-information powers, once granted, tighten compliance sharply, and that the July delay reads in hindsight as a detour rather than a derailment. Skeptics who declare the policy dead are getting ahead of the evidence just as its architects once got ahead of the enforcement.

The record supports something narrower and firmer. A world-first policy sold on the strength of five million deleted accounts has, by its own regulator's numbers, left roughly seven in ten targeted children still online, and the legislative instrument meant to close that gap is parked in committee until the inquiry reports. For any government tempted to treat Australia as a finished blueprint, the honest reading is that the blueprint is still being debugged. The correct posture toward an unproven mechanism is patience and scrutiny, not imitation. Australia agreed to run the experiment first. The least the rest of the world can do is wait for the results before signing up to repeat it.