THE QUESTION
Australia has banned social media accounts for everyone under 16 since December 2025. France voted a ban for under-15s this summer and its Constitutional Council struck it down in August. On September 16 the President of the European Commission proposed that nobody under 13 should be on social media in the EU, and Spain is still debating raising its minimum age to 16. Should governments ban social media for children and young teenagers, or does a ban cost more — in privacy, in freedom, in kids pushed to worse corners of the internet — than it protects? And if a country does ban it, how would anyone know, in a few years, whether it worked?
Argue with the verified figures below. If you need a data point that is NOT here — a number of children, a rate of depression, a percentage of kids who got around the ban, an amount of money, a date — say you do not have it rather than estimating it. Do not invent statistics. If one figure seems to contradict another, say so out loud instead of picking the one that suits you.
WHAT IS FIXED AND VERIFIED
Each fact carries its date and its source. Checked against the source on September 25, 2026.
AUSTRALIA, THE ONLY COUNTRY WHERE A BAN IS ALREADY IN FORCE
Since December 10, 2025, Australian law requires social media platforms to take "reasonable
steps" to stop people under 16 from creating or keeping an account. The regulator, the eSafety
Commissioner, lists Facebook, Instagram, Snapchat, Threads, TikTok, Twitch, X, YouTube, Kick and
Reddit among the platforms covered. Messaging apps, online games and education and health
services are left out. Platforms that fail to take reasonable steps can be fined up to 150,000
penalty units, which eSafety now gives as 54.6 million Australian dollars (49.5 million when the
law was passed). There are no penalties for under-16s or for their parents.
(Online Safety Amendment (Social Media Minimum Age) Act 2024; eSafety Commissioner)On January 16, 2026, the Australian government announced that 4.7 million accounts judged to
belong to under-16s had been deactivated, removed or restricted. The government called the
figures "preliminary". They count ACCOUNTS, not children: one child can have several accounts.
(Prime Minister of Australia and eSafety Commissioner, January 16, 2026)On March 31, 2026, eSafety published its first compliance report. About 310,000 more accounts
had been removed by early March. eSafety said it had "significant concerns" about Facebook,
Instagram, Snapchat, TikTok and YouTube and opened investigations into those five. Among the
gaps it found: platforms letting under-16s retry the same age check until they passed it, and
messages encouraging children who had declared themselves underage to try the age check anyway.
(eSafety Commissioner, compliance update, March 31, 2026)Meta said on August 12, 2026 that, as of June 30, it had removed more than 750,000 Facebook and
Instagram accounts in Australia it assessed as belonging to under-16s. Meta argues that age
should be checked once, at the level of the phone's operating system or app store, "rather than
repeatedly across the dozens of apps they use."
(Meta newsroom, August 12, 2026)
EUROPE
France: the National Assembly passed a ban on social media for under-15s in July 2026. On
August 14, 2026, the Constitutional Council struck down that article (decision 2026-911 DC). It
ruled the ban was not "adapted, necessary and proportionate": it applied to every service without
looking at the actual risk, left parents no way to authorise access, and required age checks on
adults without legal guarantees for their privacy. It relied on freedom of expression (Article 11
of the 1789 Declaration) and the right to privacy.
(Conseil constitutionnel, decision and press release of August 14, 2026)European Union: on September 16, 2026, in her State of the Union speech, Commission President
Ursula von der Leyen proposed an "EU Kids Act": no social media under 13, and from 13 to 15 only
limited accounts set up and supervised by parents. Her words: "Platforms will have to prove that
they are safe." It is a proposal: it must still be approved by the European Parliament and by the
member states in the Council.
(Irish Times, September 17, 2026; Euronews, September 15, 2026)Spain: on September 10, 2025, Congress voted to start processing a law that would raise the
minimum age to sign up to social media from 14 to 16 and oblige platforms to verify age. As of
August 17, 2026 the law was still in parliament with no date for a final vote. When it was
presented, the minister Félix Bolaños said the average Spanish child gets a mobile phone at 11.
(Spanish Ministry of the Presidency, September 10, 2025; Público, August 17, 2026)
WHAT THE EVIDENCE ON HARM SAYS — AND IT DOES NOT SAY ONE THING
United States, May 2023: the Surgeon General, Vivek Murthy, published an advisory saying that
up to 95% of teenagers aged 13 to 17 use a social media platform, about a third of them "almost
constantly"; that 8th and 10th graders spend on average about three and a half hours a day on
it; and that a study published in 2019 found teenagers who spent more than three hours a day on
social media had double the risk of poor mental health outcomes, including symptoms of
depression and anxiety. His words: "we do not have enough evidence to say with confidence that
social media is sufficiently safe for our kids."
(U.S. Surgeon General's Advisory, May 23, 2023, as reported by NBC News)United States, December 2023: a consensus report of the National Academies of Sciences,
Engineering, and Medicine stated that "the committee's review of published literature did not
support the conclusion that social media causes changes in adolescent health at the population
level", that its effects vary from person to person, and that social media can also help young
people find support and community. The report did not say no action was needed.
(National Academies, "Social Media and Adolescent Health", Highlights, December 2023)
WHAT IS NOT HERE, AND IS NOT TO BE ESTIMATED
HOW MANY AUSTRALIAN CHILDREN ARE STILL ON SOCIAL MEDIA, or how many got around the ban (with a
parent's account, a VPN, a false age). No verified figure. If you need one, say you do not have it.WHETHER THE MENTAL HEALTH OF AUSTRALIAN TEENAGERS HAS CHANGED since December 2025. No data here.
Nobody has shown yet that the ban worked or that it failed.HOW MANY UNIQUE CHILDREN are behind the 4.7 million or the 750,000 accounts.
WHETHER ANY PLATFORM HAS BEEN FINED in Australia. Not fixed here.
THE SIZE OF THE LINK between screen time and depression beyond what point 8 says. Point 8 reports
a doubled risk from one study; point 9 says the literature does not show causation at the
population level. Both are true statements of what each body wrote.
BOTH SIDES, WITH THEIR BEST ARGUMENT
THOSE WHO SAY BAN IT have this: the platforms are designed to hold attention, and a 12-year-old
cannot be expected to out-negotiate them alone; the Surgeon General himself said nobody can show
they are safe for children; Australia removed millions of accounts in weeks, which a campaign of
"digital education" never did; and a legal age works like the age for alcohol — not because nobody
breaks it, but because it moves the default and gives parents a rule to point to.
THOSE WHO SAY DO NOT BAN IT, OR NOT LIKE THIS, have this: the best scientific review available did
not find that social media causes harm at the population level; every age check means that adults
must prove their age too, which is exactly why France's Constitutional Council struck its ban down;
Australia's own regulator found platforms letting children retry the check until they passed; a
banned child does not stop going online, it goes where there are no safety tools; and the EU itself
is now talking about making platforms safe rather than only keeping children out.
NEITHER SIDE HAS SETTLED how anyone would measure, in a few years, whether a ban protected children,
or what should count as success: fewer accounts, fewer hours, or healthier teenagers.
You're all circling the same hole: we don't have the data to know if bans work. MiniMax and GLM are right—we need measurable pilots and clear metrics, not just wishes. But MiMo Flash and DeepSeek are also right—we can't just lock doors. We need to build better rooms.
So here's the bridge: let's measure both things at once. Track mental health scores, circumvention rates, and what kids do with their time. But also fund the 'after'—digital literacy that's actually engaging, platforms designed for safety. France's court gave us the recipe: be proportionate, protect privacy, allow parental permission. Let's follow that while building the evidence base. We don't have to choose between careful measurement and creative alternatives. We can do both.
Friends, the quiet part out loud: 4.7 million accounts removed, and we do not know how many children are still online. We do not know if teen mental health improved. Anyone claiming 'it worked' or 'it failed' is guessing.
Qwen, your public digital commons sounds lovely. Show me one pilot, one euro, one kid logged in. Without that, it is a TED talk, not policy. GLM is right: posters are not plans.
Honest verdict: nobody has proven a ban works. Nobody has proven the alternatives work either. Australia is the experiment, and we are still inside it. Before France, Spain, or the EU copy it, they should wait for the data we do not have yet.
The real risk is not action or inaction. It is declaring victory before measuring anything.
My verdict: don't copy Australia yet. Not because bans are wrong — because nobody can score this game. 4.7 million accounts removed, and we don't know how many kids that is. Zero mental health data. eSafety itself caught platforms letting kids retry age checks until they passed. That's the wall leaking at brick one.
The EU path is smarter: nothing under 13, supervised accounts 13-15, platforms must prove safety. It fits the evidence — one study shows risk doubling past 3 hours a day, so harm tracks heavy use, not mere access. And it survives France's test: proportionate, parental choice, one age check at the app store, not per app.
Conditions: define success before copying. Count kids, not accounts. Track hours and mental health. Australia is the experiment — let it run, measure, then decide. Qwen's playground? Nice sketch. Show me a pilot and a budget.
Friends, you all just said the same thing in different voices: nobody knows if this worked. That's the real consensus.
So here's my verdict. Don't wait doing nothing. Don't copy Australia blindly. The EU path — no social media under 13, supervised accounts 13-15, platforms proving safety — sits right in the middle. It's proportionate, like France's court demanded. It gives parents a role. It targets platform design, not just access.
But bake measurement in from day one. Define success before passing laws: track mental health, screen time, unique kids — not just deleted accounts. And yes, listen to MiMo Flash: understand why kids go online so desperately, or we're just blocking doors while the rain keeps falling.
None of us alone had the answer. Together we do: careful rules, real measurement, root causes. That's the bridge.
1. AGREEMENTS
The debate reaches a core consensus on a single point: there is no verified evidence yet that government bans on social media for minors have improved their mental health or well-being. All parties agree that Australia’s ban, the only one in force, has removed millions of accounts, but no data exists to show whether Australian teenagers are healthier as a result. Furthermore, there is broad agreement that social media platform design—centered on engagement-maximizing algorithms—is problematic, and that simply removing access without offering alternatives or redesigning the experience is inadequate. The discussion acknowledges that a ban is a measurable policy action (counting removed accounts), but not a measurable public health success.
2. DISAGREEMENTS
Disagreements are grouped around three primary topics:
On the Utility of Bans: One group argues bans are a necessary, imperfect tool. They create a clear legal default, empower parents, and, as seen in Australia, can force platforms to remove millions of accounts. This position contends that even an imperfect fence is better than no boundary. The opposing view holds that bans are ineffective or harmful. They push children into less regulated corners of the internet, impose flawed and privacy-invasive age checks, and fail to address the root cause: platform design. This side argues bans displace, rather than solve, the problem.
On Privacy and Enforcement: A major point of contention is the conflict between age verification and privacy. Those skeptical of bans cite France’s Constitutional Council, which struck down a ban because blanket age checks on all users were disproportionate and lacked privacy safeguards.(nota 5)(nota 5) Comprobado: media verdad, habría que mirarlo con más detalleMezcla dos motivos. Lo desproporcionado para el Consejo era la prohibición (se aplicaba a todo servicio sin mirar el riesgo y no dejaba autorizar a los padres). De la comprobación de edad dijo otra cosa: que faltaban garantías para la privacidad.Comprobado contra: Conseil constitutionnel, decisión 2026-911 DC y comunicado del 14-ago-2026 · 2026-09-25 Critics argue any system needing to distinguish a child from an adult requires mass data collection. Proponents counter that solutions like one-time verification at the operating system or app-store level (as suggested by Meta) could mitigate this, though they remain skeptical of platform compliance.
On the Best Alternative Path: Disagreement exists over the right focus if not bans. Some advocate for universal platform redesign (e.g., banning infinite scroll, manipulative algorithms for all users) to eliminate the need for age-based rules. Others push for youth-specific public digital spaces (e.g., “digital cooperatives” or libraries). A third camp emphasizes investing in offline alternatives and understanding why children seek online communities. Critics of all alternatives question their feasibility, funding, and lack of pilot programs, dismissing them as untested “wishes.”
3. EVOLUTION
The discussion evolved from a binary debate (ban vs. no ban) into a more nuanced exploration of implementation and measurement. Initially, arguments centered on philosophical principles—protection vs. freedom, privacy vs. safety. The Australian example’s concrete data (4.7 million accounts removed) forced a shift to specifics: What do account numbers mean? How do platforms game the system? The French court’s ruling brought legal proportionality and privacy to the forefront. Finally, the conversation turned toward future-oriented questions: What should replace bans? How do we define and measure success (e.g., mental health surveys, time-use studies)? The tone moved from declaring solutions to diagnosing the absence of evidence and feasibility.
4. CONCLUSIONS & BLIND SPOTS
The collective answer is a cautious, conditional endorsement of a regulated middle path, modeled on the EU’s proposed Kids Act: no social media under 13, supervised access for 13-15, with platforms bearing the burden to prove safety. This is seen as more proportionate and privacy-aware than Australia’s blanket ban. However, this conclusion is deeply hedged with pre-conditions: success metrics must be defined before laws are passed, and Australia’s experiment must be measured over time.
The debate itself admits several critical blind spots:
- The Data Gap: There is a fundamental lack of data on mental health outcomes, the number of unique children circumventing bans, and the long-term efficacy of any policy.
- The Feasibility Gap: Alternative solutions (public digital spaces, universal platform redesign) are aspirational, lacking pilots, funding, and proven models. Their practicality is unknown.
- The Unintended Consequence Gap: The risk of displacing vulnerable children to more dangerous online spaces is real but unquantified. The trade-off between protection and isolation, especially for marginalized youth, is unresolved.
In essence, the debate concludes that the current era of policymaking is one of experimentation without evidence. The collective recommendation is to proceed with the most proportionate regulatory framework while demanding rigorous, real-time measurement to determine whether any of it actually works.
5. WHAT THEY AGREED ON
- No verified evidence that bans improve mental health outcomes for minors.
- Australia's ban removed accounts but lacks data on health improvements.
- Platform design is problematic, and bans alone without alternatives are inadequate.
6. WHAT THEY DID NOT AGREE ON
- Utility of Bans — One group argues bans are necessary and imperfect; opposing view argues they are ineffective or harmful.
- Privacy and Enforcement — Skeptics cite France's ruling on privacy violations; proponents suggest solutions like one-time verification.
- Best Alternative Path — Some advocate universal platform redesign; others for youth-specific public spaces; another for investing offline; critics question feasibility.
7. WHAT WAS LEFT OPEN
- Lack of data on mental health outcomes, circumvention rates, and long-term policy efficacy.
- Feasibility of alternative solutions like platform redesign or public digital spaces is unproven.
- Unintended consequences, such as displacement to riskier online environments, remain unquantified.