Australia's High Court is now hearing Reddit and a libertarian advocacy group argue that the under-16 social media ban is invalid. In the same weeks, the government has raised the ceiling on penalties for breaching it. The two developments show the weakness in the policy: the state is escalating enforcement of a law whose design is under constitutional attack.
The case for the ban, stated fairly
The government's argument is serious. Parents face collective-action problems, because one child leaving a platform just isolates that child. Platforms' engagement systems are built by large teams optimising for attention, and children are the least able to resist them. Supporters say that a clear statutory floor, enforced against companies rather than families, is simpler than asking millions of households to police this individually. The government says the results so far support it: its June 2026 announcement claims more than 5 million under-16 accounts have been removed, deactivated or restricted since the ban began on 10 December 2025.
What the law actually does
The Online Safety Amendment (Social Media Minimum Age) Act 2024 was assented to on 10 December 2024. It inserted a duty into the Online Safety Act 2021: under section 63D, a provider of an age-restricted social media platform "must take reasonable steps to prevent age-restricted users having accounts." The civil penalty is 30,000 penalty units, which the government put at up to A$49.5 million for systemic breaches when the bill passed on 29 November 2024.
The duty is attached to a category of user, not to any specific harm. A teenager who wants to read a political subreddit, follow a candidate or join a community of people with the same illness is treated the same as one doomscrolling short video. That is the point on which Reddit's challenge concentrates.
The constitutional argument is about design
As Reuters reported, Reddit's lawyer Perry Herzfeld called the law "ill thought out" and "inept" and described it as "a very profound restriction on political debate in Australia." He also said: "If the Commonwealth is worried about particular features (of social media), it should regulate the features, which it has not done." The government has said it will defend the law.
That framing matters. Australia's High Court has long recognised an implied freedom of political communication. Laws that burden it must be justified as reasonably appropriate and adapted to a legitimate end. The test, in effect, asks whether a less restrictive route was available. Herzfeld's point is that one plainly was. A duty aimed at specific design features such as infinite scroll, autoplay, algorithmic amplification of strangers' content and unsolicited contact would address the harms the government cites without cutting off an entire age group from political discussion. I can't predict how the Court will rule. But the proportionality question is a real one, and a regulator that skipped the narrower option has to answer it.
Escalation is not a substitute for fit
On 28 June 2026 the Prime Minister and Communications Minister Anika Wells announced they would double the maximum penalty from A$49.5 million to A$99 million, "in line with penalties available under competition and consumer law." The cap is now 300,000 penalty units, which eSafety says is currently about A$109.2 million. The same release doubles the penalty for ignoring the eSafety Commissioner's information-gathering notices and widens the Commissioner's power to compel evidence, including from third parties such as age-assurance providers and app stores. It also says eSafety is investigating five platforms: Facebook, Instagram, Snapchat, TikTok and YouTube. Reuters adds that the government is preparing an enforcement lawsuit against some of the biggest platforms.
There are three problems with this.
- Reasonable steps is a moving target. A duty to take "reasonable steps" is vague by design. Doubling the penalty for a standard that platforms cannot fully predict pushes them to over-comply, meaning blunt age checks and collateral removal of lawful adult and teen speech.
- A big headline figure rewards the wrong behaviour. For the largest platforms, even A$99 million is small relative to revenue. The people it affects are smaller services, forums and non-commercial communities, which can least afford legal uncertainty. Reddit's argument that it should not be treated as conventional social media is that problem in miniature.
- Success is measured in accounts, not in outcomes. "5 million accounts restricted" counts enforcement activity. It does not show that children are safer, that they have not moved to unregulated services, or that political speech has not been chilled. The Act itself requires an independent review within two years of section 63D taking effect, which is the right place to test those claims. Penalties should not be raised before that evidence exists.
What proportionate policy would look like
A proportionate approach would keep what is defensible: strong default protections for minors, a duty on platforms to mitigate foreseeable risks to children, and real transparency about how recommendation systems treat young users. It would regulate features and systems, with penalties scaled to the size of the service and the seriousness of the failure, and would leave a pathway for teenagers to take part in civic life online with appropriate safeguards.
Whatever the High Court decides, the government's choice to respond to a legitimacy challenge with a larger fine tells us how it sees the problem. If the Court finds that the law burdens political communication more than necessary, a bigger penalty will not help it. The government would be better off using the review period to design a narrower law that does not depend on one.