Op-ed: kids dodging social media age bans shows the bans' limits
MIT Technology Review published an opinion essay by Michal Luria, Aliya Bhatia and Danah Boyd on social media age bans. It is an argument, not a news report.
The authors start from the pile-up of bans. In July, France joined Australia in restricting minors from social media, barring children under 15 from major platforms, before the ban was blocked by the country's top court on human rights grounds. The UK aims to forbid kids under 16 from Facebook, TikTok and other platforms in early 2027, and the EU has announced a plan to prohibit social media for kids under 15.
The early lessons from Australia, which enacted its ban in 2024, are that such restrictions are easily circumvented. The authors cite a study published earlier this year that found four in five teens under 16 made their way onto barred platforms. Tactics range from using a sibling's face for ID to drawing fake facial hair. Some proponents of bans want enforcement that blocks these workarounds. The authors reply that circumvention is a signal, not a nuisance or something to be simply deterred.
Their case has several strands. First, bans suppress kids' access to information that matters for development, identity formation and socialization. When young people get around restrictions, the authors say, they are often doing what we want developing minds to do: being curious, solving problems and exercising independence. Second, parents are part of the picture. When the authors examined social media safety interventions last year, parents told them they often join their kids in lies that help them reach needed but gated services. Research from 2011 showed parents teaching their children to evade Facebook age restrictions. The authors add that parents are not a monolithic group; many see responsible access as a way to help kids connect with like-minded peers, find information and learn digital safety skills early.
Third, the authors say further restriction is likely to backfire. Kids who cannot get into the platforms they want will likely move to alternative spaces, often with fewer guardrails. Some end up on harmless if unconventional venues, such as Google Docs or podcast comment sections; others on more dangerous platforms where they are vulnerable to abuse and predation. Preliminary research after the Australian ban shows kids' screen time has not meaningfully decreased but has sometimes been redirected, with some young users going to spaces unmonitored and unknown to parents. The authors argue the riskiest platforms are those with the fewest eyes on the digital street, and that outdated, dying and under-resourced platforms attract both vulnerable youth and those who prey on them.
Fourth, age gates cost everyone privacy and security. Some platforms track users' activity to confirm age, which profiles them and raises the risk of misuse or leaks. Others ask for documents such as driver's licenses, which can be mishandled or lead to identity theft. As an example, the authors say Discord reported last year that one of its third-party services, 5CA, used to confirm age, had been compromised, exposing at least 70,000 images of users' government IDs. They warn such data could be linked to accounts and private activity, revealing anything from political opinions to sexual preferences.
Efforts to stop kids getting around gates have also led policymakers to target VPNs. State age verification bills in Michigan and Utah have tried to interfere with VPN privacy protections (the Utah bill is now blocked by a federal court), and Australia's eSafety Commissioner is nudging platforms to detect and block VPN use. The authors say requiring adult users to prove their age to use a VPN defeats the purpose of the technology and can endanger journalists and human rights defenders working on sensitive issues.
What do they propose instead? Policymakers should focus on the incentives that drive companies to create risks for kids, for example by limiting the collection and use of data that companies use to target content and features to all users, including children. The authors say nearly all the parents and teens they spoke with want such restrictions, and that both groups are frustrated by features designed to keep people online longer than they want. The other approach is more user-controlled tools for families. Their example is keyword-based content filtering, which would let a student targeted with a particular nickname or inside joke block any post or comment containing it. Platforms are unlikely to know which term a child wants to block, and policy cannot address such dynamic harms directly, but it can offer tools that put power back in users' hands.
The authors close with their verdict: circumvention shows where policy falls short, and stricter enforcement will mean yet more data collection without meeting kids and parents where they are.
Key facts
- The authors cite a study published earlier this year finding that four in five teens under 16 made their way onto platforms barred under Australia's 2024 ban; tactics include using a sibling's face for ID and drawing fake facial hair.
- They argue circumvention is a signal, not a nuisance, and that closing loopholes could push kids toward less-moderated, higher-risk corners of the web while forcing everyone to hand over sensitive data.
- Their example of the data risk: Discord reported that its age-confirmation service 5CA was compromised, exposing at least 70,000 images of users' government IDs.
- Policies aimed at VPNs, including bills in Michigan and Utah (the Utah bill is now blocked by a federal court), are criticised as undermining privacy for users such as journalists and human rights defenders.
- Instead of bans, they urge limits on company data collection and use, and more user-controlled tools such as keyword-based content filtering.
Why it matters
Age bans are spreading. France restricted under-15s in July before its top court blocked the ban, the UK aims to bar under-16s in early 2027, and the EU has announced a plan for under-15s. Australia, which enacted its ban in 2024, is the first test case, and the authors read its early results as evidence that gates are easy to get past. Their larger point is about direction: tightening enforcement means more identity checks and more data collection, so the fix for leaky bans creates its own privacy problem.
Who it affects
Kids and teens, who the authors say lose access to information and social connection and may be pushed to less-monitored spaces. Parents, some of whom the authors say help their children get around age gates. Platforms, which would face either enforcement duties or limits on data collection under the authors' alternative. Adult users, who would have to submit personal information to prove their age. VPN users, particularly journalists and human rights defenders, who the authors say could face surveillance risks if VPN access is tied to age checks.
How to use it
This is an argument to weigh, not a product or tool. Read it as a framework for judging age-ban proposals: ask how circumvention would be handled, what data enforcement would require, and where displaced kids would go. The authors offer two alternatives. One is limiting the collection and use of data that companies use to target content and features. The other is expanding user-controlled tools for families, such as keyword-based content filtering that lets a student block any post or comment containing a specific nickname or inside joke.
How solid is it
This is an opinion essay by Michal Luria and Aliya Bhatia, who are at the Center for Democracy & Technology (a senior research fellow and a senior policy analyst), and Danah Boyd, the Geri Gay Professor of Communication at Cornell University. Its factual anchors are a study of Australian teens (four in five under 16 got onto barred platforms), preliminary research on screen time after the ban, 2011 research on parents helping kids evade Facebook age limits, and the authors' own work with parents and teens. The piece does not name the study behind the four-in-five figure and does not describe the methodology or sample of the authors' parent and teen findings. It includes no responses from governments or proponents of the bans.
Risks and caveats
The central prediction, that further restriction is likely to backfire, is the authors' judgment rather than a demonstrated result. The evidence on displacement is described as preliminary, and the piece gives no figures on how much kids' screen time changed in Australia. The Discord breach is an illustration of data risk; the source does not say it was caused by or tied to the Australian ban. The authors' alternatives (data limits, filtering tools) are recommendations, and the essay does not show they would protect kids better than bans do.
“circumvention is a signal, not a nuisance or something to be simply deterred”
— Michal Luria, Aliya Bhatia and Danah Boyd, MIT Technology Review