At What Age Should Kids Use Social Media? France, Australia, and the Research

In the space of about seven months, two democracies did something no country had tried before. In December 2025, Australia's ban on social media for anyone under 16 took effect. In July 2026, France's National Assembly gave final approval to a ban for under-15s, the first of its kind in the European Union. Several more countries are lined up behind them.

This is not a piece about whether you are a good parent for letting your kid on Instagram, or a bad one for not. That framing is useless, and it is not what any of the research is actually about. This is about the evidence underneath the headlines: what technology does to kids at different ages, what genuinely helps, what genuinely hurts, and what is fine as long as someone is paying attention. The new laws are just a good excuse to look at it clearly.

Here is what happened, and what the science underneath it says.

What Australia and France actually did

Australia went first. Its Online Safety Amendment (Social Media Minimum Age) Act took effect on 10 December 2025 and sets a minimum age of 16 to hold an account on ten named platforms, among them Facebook, Instagram, TikTok, Snapchat, YouTube, and Reddit. Messaging, gaming, and education tools like WhatsApp, Discord, Roblox, and YouTube Kids are exempt. One design choice matters more than the age itself: the law puts the burden on the platforms, not families. Companies that fail to take reasonable steps face fines. There is no penalty for a child or a parent.

France set its line at 15. Its ban, approved on 21 July 2026, is the first in the EU, but it is worth being precise about its status, because the headlines ran ahead of the reality. The law has been adopted, not switched on. To avoid clashing with EU law, lawmakers removed the actual enforcement machinery from the final text, the European Commission has flagged it as touching EU authority, and a constitutional review is pending. Legal analysts expect it cannot realistically take effect on its planned September 2026 date. For now it is closer to a stated intention than a working rule.

The bigger pattern

These two are the visible edge of a wave. In November 2025 the European Parliament voted for a non-binding EU-wide minimum age of 16. Denmark, Greece, Norway, and Spain are each moving toward their own limits. In the United States, Florida's law barring accounts for under-14s is now being enforced while similar laws elsewhere sit in court.

One fact reframes the entire debate. The familiar American cutoff of 13, the age most platforms ask you to be, was never a developmental finding. It comes from COPPA, a 1998 privacy law about collecting data from children. Thirteen was a line drawn around data, not around readiness. These new 15-and-16 laws are the first serious attempt to set the age on health and development instead. Whether they picked the right number is exactly the open question.

Does the research back the bans?

Two honest answers, because the evidence genuinely supports both concern and humility.

The worry is real and official. The 2023 US Surgeon General's advisory did not call social media proven harmful. It said we 'do not yet have enough evidence to determine if social media is sufficiently safe,' and called adolescence 'a particularly vulnerable period of brain development.' It pointed to research that teens who spend more than three hours a day on social media face roughly double the risk of depression and anxiety symptoms, from a study of 6,595 US adolescents. Since up to 95% of 13-to-17-year-olds use social media, that is a lot of kids in the exposed group.

The counterweight is just as serious. A 2019 analysis of more than 355,000 adolescents found that technology use explained less than half a percent of the difference in their well-being. A 2024 review in Nature, responding to Jonathan Haidt's widely read book 'The Anxious Generation,' argued plainly that 'there is no evidence that using these platforms is rewiring children's brains or driving an epidemic of mental illness.' And the National Academies of Sciences concluded in 2023 that the evidence did not support the claim that social media causes population-level changes in adolescent health, while stressing, in the same breath, that it has not been shown to be safe.

Both are true. The correlations are real and the causation is unproven. The researchers who spend their careers on this disagree, in good faith, about how much to make of the same numbers.

Australia's early experience shows a different limit. A study in The BMJ in June 2026 found that in the first three months of the ban, more than 85% of under-16s were still using restricted platforms, through their own, borrowed, or fake accounts. A law can draw a line. It cannot make a motivated teenager honor it.

The part that is actually settled

Here is the useful news hiding under the argument about bans. While the causation fight continues, pediatric authorities broadly agree on how technology should scale with age. This is the part a parent can act on today, no matter what any government does.

The most important shift is in the framing. The American Academy of Pediatrics' 2026 policy stepped back from a single universal hour limit and toward three things that matter more: the quality of the content, the context it is used in, and whether an adult is involved. Its plain-language framework for parents is a set of five Cs: Child, Content, Calm, Crowding Out, and Communication. The question stopped being only how long, and became what, and with whom.

Roughly by age, the guidance lands here:

• Under 18 months: screens are discouraged except for video chat. Video calling a grandparent counts as an exception because it is interactive and a real person answers back. The World Health Organization goes further, advising no screen time at all before age two.
• Ages 2 to 5: up to about an hour a day of high-quality programming, watched together. The 'together' is not filler. A 2020 study of 488 children found that when a parent stayed engaged alongside the child, the negative link between media and early literacy vanished. Same screen, opposite result, because of who was sitting next to them.
• Ages 6 to 12: consistent, individual limits, and one rule that does most of the work, that screens do not crowd out sleep, movement, and time with people. The AAP specifically recommends keeping devices out of bedrooms and off about an hour before bed. Its free Family Media Plan is a reasonable starting point.
• Teens: this is where social media arrives, and where the evidence is genuinely mixed. Some of the risk looks age-specific. A 2022 study in Nature Communications found the negative link between social media and life satisfaction was sharpest in narrow windows, around ages 11 to 13 for girls and 14 to 15 for boys. But teens also report real gains. In Pew's 2025 survey, 74% said social media makes them feel more connected to their friends and 63% called it a place to be creative. For LGBTQ+ teens in particular, having even one online space where they feel safe is tied to lower anxiety and lower suicide risk.

The clearest harms are worth stating just as plainly. Heavy, unsupervised screen use in the youngest years is linked to delays in language and thinking. And at every age, the most consistent damage is to sleep. In Pew's data, 45% of teens said social media hurts their sleep, and among teen girls it was half. If you change one thing, make it the phone that charges outside the bedroom. We went deep on screen time rules for teens that actually hold if that is the fight in your house.

The through-line nobody can legislate

Notice what runs through all of it. It is not screen time. It is involvement.

What the research keeps rewarding is not a stricter number or a later birthday. It is an adult in the loop: watching alongside a toddler, choosing the content, keeping the phone out of the bedroom, staying in conversation with a teenager about what they are running into. The AAP's 2026 guidance says outright that a strong caregiver relationship is itself protective.

That is not a call for more vigilance from already-stretched parents, and it is not a judgment on anyone who hands over a tablet to get through a hard afternoon. Every parent does that. It is simply where the leverage is. A law sets one line for a whole country. The evidence says the outcome turns on the specific app, the specific content, and whether anyone is nearby, which happen to be the things a family can actually adjust.

Australia and France picked a number. You do not have to wait for a law, or agree with one, to use what the research already shows: that 'okay for a kid' was never a single age, and never really about the clock.

Where to start

That last idea, that 'okay' depends on the specific thing and not a blanket age, is the whole reason we built the Yulixis decoder. It takes the games and apps kids actually beg for and gives each one a straight answer: the real age it fits, the few things that genuinely matter, and the one setting worth changing. It is free, we add new titles every month, and getting it puts you on our email list. If you have ever stood in the kitchen at 10pm trying to decide whether to say yes, that is exactly what it is for.