The ban and its central contradiction
Australia’s under 16 social media law took effect in December 2025. Under the law, age restricted social media platforms must take reasonable steps to stop Australians under 16 from creating or keeping accounts. The government says the law responds to harms linked to algorithmic recommendation, infinite scroll, auto play, and other features that keep young users online for longer and expose them to harmful material (Australian Government, 2026; eSafety Commissioner, 2026). Public support for the law makes sense. Many parents are worried about addictive feeds, sexualised content, self harm content, bullying, and the pressure social media puts on children every day. The law speaks directly to those fears. At the same time, the policy directs most of its force toward children’s access to accounts. The systems that rank content, shape visibility, and reward constant engagement continue to operate in much the same way. This creates a basic policy gap. The law recognises that platform design can harm children, yet its main solution controls entry to the platform, while the design logic inside the platform stays largely in place.
The strongest part of the law is also the part that reveals its limit. Official fact sheets describe age restricted platforms as services that use recommender features, infinite scrolling, visible counts such as likes, and disappearing content such as stories (Department of Infrastructure, Transport, Regional Development, Communications, Sport and the Arts, 2026). It does not treat harm as a simple matter of individual weakness. Harm is built into the environment. The term algorithmic power is used to explain how software shapes what becomes visible and what stays hidden (Bucher, 2012). On social media, children do not just choose content in a neutral space. They move through feeds that are ranked, predicted, and adjusted in real time. Recommender systems can normalise toxic material through repeated exposure, especially when increasingly intense content is pushed into young people’s feeds (Regehr et al., 2025). Once the problem is understood in those terms, the law’s narrow focus on account holding looks incomplete. The source of harm lies in a system of organised attention, not only in the age of the user.
Platform scholars have spent years arguing that digital harms cannot be understood only at the level of single pieces of content. Platform power works through ecosystems, business models, governance rules, and technical infrastructures (Van Dijck, Poell, Nieborg, 2019). Public debate often narrows platform problems into manageable topics while leaving the deeper structure of the platforms untouched (Gillespie, 2023). Australia’s law fits that pattern. It acknowledges harmful design, yet the practical duty imposed on companies is to identify and remove younger users from account based spaces. There is little equivalent pressure on platforms to reduce recommendation intensity, slow down engagement loops, remove visible popularity metrics, or redesign feeds for safety by default. A child may be blocked from opening an account, while the same platform still runs on the same attention economy for everyone else. The policy changes the border of the system far more than the system itself.
What this law really reveals is a familiar habit in digital governance. When a technology problem becomes politically urgent, governments often reach for the part that is easiest to measure. In this case, that means age, account access, and removal numbers. These are visible outcomes. They can be counted, reported, and turned into proof that action is happening. The harder questions stay in the background. What kind of platform design keeps young people trapped in cycles of comparison and emotional dependence? Why are business models built around keeping users overstimulated for as long as possible? Those questions are much less convenient because they do not lead to quick policy wins. They lead to conflict with powerful companies. That is why this ban feels clear on the surface but limited underneath. Once the debate is framed around whether children should be allowed onto platforms, it becomes easier to forget that regulation could aim higher. It could ask platforms to change the emotional tone of feeds, reduce compulsive design, or slow down the spread of harmful content. Instead, the public gets used to a smaller ambition. Safety starts to mean restricting access, rather than transforming the environment. That shift matters because it changes what people imagine is possible. Over time, a narrow policy can also produce a narrow political imagination.
Why set an age line can not solve the problem
Research on children’s online risk also suggests that a single age threshold cannot solve a layered problem. Livingstone and Stoilova (2021) classify online risk through four broad categories that are content, contact, conduct, and contract. This framework shows that children face many kinds of risk at the same time. Some risks come from what they see. Some come from who reaches them. Some grow out of peer behaviour. Some are tied to data collection and commercial pressure. A similar conclusion in their review of online safety research, showing that children’s digital risk is social, technical, and economic all at once (Jang and Ko, 2023). A minimum age rule may reduce exposure for some children, especially younger ones. Yet it does very little to change the way recommendation systems sort content, the way metrics fuel comparison, or the way commercial design keeps attention circulating. The law simplifies a broad risk environment into one gatekeeping rule. That makes enforcement easier to explain, though it does not make the deeper problem easier to solve.
The law also depends on data driven classification. Government fact sheets explain that platforms may use age assurance technologies, including age inference, to work out whether a user is over 16, and such systems may rely on signals such as account history, interests, networks, or other available information (Australian Government, 2026). The Office of the Australian Information Commissioner also notes that compliance with the minimum age rules will involve the handling of personal information about whether a person ordinarily lives in Australia and whether they are over 16 (Australian Information Commissioner, 2025). This means the policy expands a familiar digital logic. Children are cast as people who need protection, while they are also turned into data subjects whose age must be estimated, checked, or inferred. Classification becomes part of care. Sorting becomes part of safety. That move may be administratively practical, yet it also shows how easily child protection can slide into more data collection and more platform judgment about users.
Social media should not become a convenient scapegoat for wider mental health problems (Christensen et al., 2024). Their point is not that social media is harmless. Their point is that youth distress has many causes, and policy can become distorted when one cause is treated as the whole explanation.
The authentication method
Early implementation shows how hard it is to govern algorithmic environments through access control alone. In its March 2026 compliance update, eSafety reported that platforms had told the regulator about 4.7 million removed or restricted age restricted accounts by mid December 2025, with more than 300,000 additional accounts prevented from signing in by early March 2026 (eSafety Commissioner, 2026, March ). But many children kept their accounts because some platforms had not asked them to verify their age, while others were still able to get around weak checks and stay active online. A 15 years old teenager named Noah Jones, who continued using social media after Australia’s under 16 law came into effect in December 2025. According to The Guardian (Taylor, 2026), Noah said he was still able to stay on the platforms with little disruption, even after millions of accounts had been removed. The same report said more than two-thirds of teens were still active on restricted platforms because some age checks were weak and easy to get around. A 15 years old boy in Sydney also passed Snapchat’s visual age check and kept access to the app (Taylor, 2025). The system used facial age estimation, but it judged him as older than 16. The real difficulty begins when platforms try to enforce that rule across millions of users. If a 15 year old teenager can stay online with limited disruption, the ban starts to look less like a firm barrier and more like a patchy filter. It also shows that children do not simply accept regulation in a passive way. They test it, work around it, and keep using digital media that are still open to them. And if platforms rely on age estimation tools, those tools can still make mistakes. Once that happens, enforcement depends on technology that looks precise but does not always work in a precise way.
Children are presented mainly as people who need to be removed from danger, rather than as users moving through systems that adults have already failed to shape responsibly. Parents may be told that the state has acted. Platforms may claim they have complied. Politicians may point to enforcement numbers. All of that creates a sense that the issue is being brought under control. But the everyday pressures shaping young people’s digital lives do not disappear so neatly. The search for approval, the pressure to perform, the pull of endless content, and the fear of being left out can move across different apps and spaces very quickly. A policy built around one legal threshold can look solid while daily digital pressure remains fluid, portable, and deeply embedded in ordinary life.
The bans have practical limits, can be bypassed, and may push young people toward other digital spaces that carry different risks(Champion et al., 2025). These arguments push the conversation away from a fantasy of total protection. Children live in digital cultures that stretch across messaging apps, games, video platforms, and social networks. One legal threshold on account holding cannot redesign that culture on its own. The Australian law offers a strong political gesture, though gestures and structural reform are not the same thing.
A more effective response would put more pressure on platform design itself. Chhabra et al. (2025) reviewed recommendations on social media and youth mental health and found strong support for transparency, safer design, and stronger accountability for companies. That direction points to practical reforms. Platforms could be required to reduce harmful amplification for younger users, remove default infinite feeds, limit visible popularity metrics, and open their systems to independent auditing. Safety could be built into recommendation design instead of being left at the gate. Australia is already moving toward further regulation, with debate growing around a digital duty of care and wider platform responsibilities. That shift would go closer to the source of the problem. A child safety policy becomes more convincing when it changes how platforms work, not only who can enter them.
The ban also shapes the kind of internet young people are being prepared to enter. If digital harm is managed mainly by delaying access, the policy buys time, but it does not change the space that young people will eventually move into. The age of sixteen gives the law a clear line, and that line is easy to defend in public. Yet children do not suddenly become digitally mature on one birthday. Some younger teenagers may already be careful, supported, and capable of managing online risks. Some older teenagers may still struggle with pressure, manipulation, or harmful comparison. This matters because the law turns a complex issue into a simple rule. It sets an age threshold, while the same platforms, the same design features, and the same data driven systems remain in place. From a digital rights perspective, the problem is not only when young people are allowed to enter online spaces. The problem is also what kind of spaces they are being allowed to enter, and what rights, protections, and control they will actually have once they get there. Delaying access can look protective, but delay on its own does not make the future internet safer or fairer.
References
- Australian Government. (2026, January). Social media minimum age fact sheet. Department of Infrastructure, Transport, Regional Development, Communications, Sport and the Arts.
- Bucher, T. (2012). Want to be on the top? Algorithmic power and the threat of invisibility on Facebook. New Media & Society, 14(7), 1164–1180. https://doi.org/10.1177/1461444812440159
- Chhabra, J., Pilkington, V., Benakovic, R., Wilson, M., Sala, L.L. and Seidler, Z. (2025). Social media and youth mental health: a scoping review of platform and policy recommendations (Preprint). Journal of Medical Internet Research, 27. doi:https://doi.org/10.2196/72061.
- Christensen, H., Slade, A., & Whitton, A. E. (2024). Social media: The root cause of rising youth self-harm or a convenient scapegoat? Medical Journal of Australia, 221(10), 524–526. https://doi.org/10.5694/mja2.52503
- Department of Infrastructure, Transport, Regional Development, Communications, Sport and the Arts. (2026). Social media minimum age. Australian Government.
- eSafety Commissioner. (2026). Social media age restrictions.
- eSafety Commissioner. (2026, March). Social media minimum age: Compliance update.
- Champion, K.E., Birrell, L., Smout, S., Maree Teesson and Slade, T. (2025). Debate: Social media in children and young people – time for a ban? Beyond the ban – empowering parents and schools to keep adolescents safe on social media. Child and Adolescent Mental Health. doi:https://doi.org/10.1111/camh.70032.
- Gillespie, T. (2023). The fact of content moderation; or, let’s not solve the platforms’ problems for them. Media and Communication, 11(2), 406–409. https://doi.org/10.17645/mac.v11i2.6610
- Jang, Y. and Ko, B. (2023). Online Safety for Children and Youth under the 4Cs Framework—A Focus on Digital Policies in Australia, Canada, and the UK. Children (Basel), [online] 10(8), pp.1415–1415. doi:https://doi.org/10.3390/children10081415.
- Livingstone, S. and Stoilova, M. (2021). The 4Cs: Classifying Online Risk to Children. The 4Cs: Classifying Online Risk to Children. doi:https://doi.org/10.21241/ssoar.71817.
- Regehr, K., Shaughnessy, C., Zhao, M., Cambazoglu, I., Turner, A. and Shaughnessy, N. (2025). Normalizing toxicity: the role of recommender algorithms for young people’s mental health and social wellbeing. Frontiers in Psychology, 16. doi:https://doi.org/10.3389/fpsyg.2025.1523649.
- Taylor, J. (2025, December 8). Boy, 15, clears Snapchat’s visual age check ahead of Australia’s social media ban taking full effect. The Guardian.
- Taylor, J. (2026, April 11). Fifteen-year-old Noah hasn’t been kicked off any social media platforms, he’s still fighting Australia’s under-16 ban in court. The Guardian.
- van Dijck, J., Poell, T., & Nieborg, D. (2019). Reframing platform power. Internet Policy Review, 8(2). https://doi.org/10.14763/2019.2.1414
Be the first to comment