The World Is Banning Kids From Social Media. Nobody Agrees on How.

Australia banned kids under 16 from social media in December. France followed in January. Karnataka — India's tech capital — announced its own ban yesterday. Norway, Denmark, and Malaysia are drafting theirs. Eight US states have passed laws. Over 300 bills are pending across 45 more.

Everyone agrees kids and social media is a problem. Nobody agrees on the fix.

The Spectrum

Here's what makes this story interesting: every country is trying something different.

Australia went hardest. Full ban for under-16s. Platforms must block new accounts, deactivate existing ones. Companies face fines up to A$49.5 million if they don't comply. Kids and parents face zero penalties — the burden sits entirely on platforms. France approved a ban for under-15s in January 2026, set to take effect September. Same idea, slightly younger cutoff. Norway is drafting a ban for under-15s. Denmark wants one by 2026. Spain, Italy, Greece, Finland, and Germany are all "considering" restrictions — which in EU speak means years of debate ahead. Virginia took a completely different route. No ban. Instead: a one-hour daily limit for under-16s, unless a parent opts in for more. Instagram, TikTok, YouTube, Snapchat — all capped at 60 minutes. Seven other US states have passed their own versions, each slightly different. China didn't ban anything. It built time limits into the operating system. Under-8s get 40 minutes of screen time per day. Ages 8-16 get one hour. Ages 16-18 get two hours. Douyin (TikTok's Chinese version) shuts off automatically. No platform choice, no workaround — the restriction lives in the device. Karnataka announced its ban yesterday during the state budget. Under-16s can't use social media. Details on enforcement haven't been released yet.

That's six approaches to the same problem, from six different places, based on six different assumptions about who's responsible.

Two Months In: How's Australia Doing?

Not great.

NPR reported in February that teenagers are bypassing age checks with ease. Face ID passes them as 16 because they look old enough. Kids who set their ages higher before the ban kept their accounts. Some platforms suspended under-16 accounts. Others didn't.

Snapchat and YouTube argued they aren't even social media companies. YouTube said the "rushed" law would actually make kids less safe — they'd still browse without accounts, losing the parental controls built into logged-in profiles.

The Guardian found that some teens feel more isolated. Others just migrated to messaging apps and gaming platforms, which aren't covered by the ban.

Australia's eSafety Commissioner calls it "a delay, not a ban." No penalties for kids or parents who get around it. The question is whether platforms will invest seriously in enforcement when the workarounds are this easy.

The Evidence Problem

Here's where it gets uncomfortable.

Nature Health published a paper last week with a blunt title: "Social media bans for teens lack evidence."

The core finding: UK schools that heavily restrict phone and social media use during school hours show no mental health differences compared to schools that don't restrict anything. Meta-analyses of trials where people quit social media found modest mental health improvements — but "modest" isn't "ban-worthy."

The concerns are real. Adolescents have less self-control. They're more vulnerable to social comparison, peer pressure, and reward signals — exactly what algorithms exploit. Exposure to harmful content (sexual material, extremism, cyberbullying, wellness scams) is well-documented.

But the leap from "this can cause harm" to "banning it fixes the harm" doesn't have strong evidence behind it. Not yet.

The researchers aren't saying "do nothing." They're saying "test it properly first." Randomized school trials with broad outcome measures. Not policy based on assumption.

The Deeper Split

Zoom out and you see two fundamentally different philosophies.

The ban model (Australia, France, Norway): social media is dangerous for kids, full stop. Remove access. Force platforms to gate-keep. The child shouldn't have to navigate the risk. The restriction model (China, Virginia, some US states): social media isn't going away. Teach kids to use it responsibly. Limit time, require parental involvement, make platforms accountable for design choices that hook young users. The EU model: study it extensively, publish guidelines, create expert panels, and maybe do something in a few years.

Each model reflects something about the culture it comes from. Australia's ban-first approach mirrors its history of internet regulation (it was one of the first countries to mandate ISP-level content filtering). China's device-level controls reflect a government comfortable building rules into infrastructure. America's patchwork of state laws reflects... well, America.

What Nobody's Measuring

Here's what's missing from almost every proposal: outcomes.

Australia didn't build evaluation into its law. France hasn't specified how it'll measure success. The US bills focus on penalties, not on tracking whether kids' mental health actually improves.

China's approach is the most measurable — screen time is quantifiable, and the government collects the data. Whether that data will ever be independently reviewed is a different question.

The one thing every expert seems to agree on: we're running a global experiment on hundreds of millions of children, and almost nobody built in a way to check whether it's working.

Where This Goes

More bans are coming. The political incentive is strong — "protect the children" polls well everywhere. The EU will eventually act. More US states will pass laws. India's other states will watch Karnataka.

The harder question is whether any of this will make kids safer, happier, or healthier. Or whether it'll just push the same problems to platforms and spaces that don't have rules yet.

The world decided social media is bad for kids. That might be right. But deciding something is true and proving it works are two different things. Right now, we're doing a lot of the first and almost none of the second.