EU Kids Act: Europe's Social Media Ban for Under-13s

If your kid is on Instagram, YouTube, Roblox or ChatGPT, the most important voice in their digital life is suddenly a German woman in Brussels. On Thursday, European Commission President Ursula von der Leyen, joined by EU tech chief Henna Virkkunen, formally presented the EU Kids Act — the most aggressive attempt any democracy has ever made to rip the engagement playbook out of Silicon Valley's hands.

The headline terms are blunt: no social media under the age of 13. No personal account under the age of 15. Children aged 13 and 14 get parent-supervised "mini accounts" — limited contacts, a screen-time cap of up to one hour a day. And for ages 15 to 18, "safe design" stops being a talking point and becomes a legal obligation: platforms will have to prove they are safe by design, with the burden of proof reversed onto the companies. "It is us who decide the rules, not big tech," von der Leyen told the European Parliament in Strasbourg on Wednesday, previewing Thursday's formal proposal.

European Commission President Ursula von der Leyen, who formally presented the Kids Act on Thursday alongside EU tech chief Henna Virkkunen. (Official 2020 portrait, European Commission; photo by Etienne Ansotte, © European Union — EU reuse license, attribution required.)

EU Kids Act rules: what each age group can and can't do

The Kids Act — which Brussels expands as "EU Keeping Internet Digital Spaces Accountable and Trustworthy" — sets up a three-tier system for young users across the bloc:

Under 13: no social media at all. Children below 13 would be banned from accessing social media services entirely.

Ages 13 and 14: parent-supervised "mini accounts." As von der Leyen put it in her State of the Union address: "That means, from 13 until a child turns 15, only mini-accounts, set up and supervised by parents or guardians, with limited features and time restriction of one hour a day." According to the Commission's Thursday statement, those safeguards include limited social contacts and a screen-time cap of up to one hour per day.

Ages 15 to 18: safe design as a platform obligation. Here the proposal flips the regulatory script. Instead of regulators proving a platform harmed a child, the platform has to prove to regulators that it is safe by design for young users. "So we are reversing the burden of proof. Platforms will have to prove to us that they are safe," von der Leyen said. "Because it is not about our minors accessing social media. It is about when and how we allow social media to access minors."

How EU Kids Act age verification would work

The enforcement machinery is the part that will keep compliance teams in Menlo Park, Mountain View and San Francisco up at night. Under the proposal, social-media and video-sharing providers must perform age checks when a user opens a new account. For existing accounts — hundreds of millions of them — providers would have to estimate a user's age from signals such as the date the account was set up or credit card details, according to the Wall Street Journal's account of the Commission's Thursday statement.

The scope deliberately reaches beyond traditional social networks. The obligations also cover video-sharing platforms, online games, and AI companions and chatbots offering services to under-18s. For young users, the proposal would require: no infinite scroll without stopping points; no algorithmic funnels drawing minors toward harmful content; profiles private by default; geolocation off by default; and built-in time-management tools. AI companions face a specific prohibition: they cannot simulate personal relationships in ways that could make a child emotionally dependent.

That last point matters more than it sounds. Conversational AI is rapidly becoming the default interface in everyday software — we've looked at how AI assistants are being built directly into the browser — and Brussels is now saying that if your chatbot feels like a friend to a child, that's a regulatory problem, not a feature.

What happens if platforms break the Kids Act?

Platforms that violate the rules risk fines of up to 6% of their global annual revenue, borrowing the Digital Services Act's penalty regime for social and video-sharing services, while AI companions and chatbots fall under the AI Act's enforcement framework. Platforms will also have to pay a supervisory fee to fund regulators' oversight of the new rules.

How we got here: leak, speech, formal proposal

The Kids Act arrived in three acts. On Monday, September 14, a draft leaked to Reuters and Bloomberg, revealing the Commission's direction. On Wednesday, September 16, von der Leyen made the Kids Act the centerpiece of her annual State of the Union address in Strasbourg — her most combative passage in the speech: "I am aware that many perceive the power of big tech as overwhelming and impossible to roll back. I disagree... Europe has the power to act." And on Thursday, September 17, she and Virkkunen formally tabled the proposal.

Only the Thursday presentation counts as the official text. Details circulating from the earlier leaked draft that did not appear in the formal announcement — enforcement funding mechanics, app-store-level verification schemes, pre-approval windows for new platform features — remain unconfirmed and should be treated as such.

The hemicycle of the European Parliament in Strasbourg, where von der Leyen unveiled the Kids Act in her State of the Union address on September 16 (plenary-session file photo, 2014). The proposal now faces years of negotiation. (Photo: David Iliff, CC BY-SA 3.0, via Wikimedia Commons.)

This is a Silicon Valley story

Make no mistake about whose products this is written for. Reuters reports that the proposal's obligations will hit Google's YouTube and Meta's Facebook and Instagram directly, and concerns cited by officials also reach TikTok and ChatGPT. The games provision puts Roblox — where American children spend enormous amounts of time — squarely in the frame. The age-verification machinery inevitably pulls in the Apple and Google app stores, through which nearly all of these services reach minors.

Von der Leyen did not pretend otherwise. At Thursday's press conference she said: "92% of Europeans say we should strengthen the protection of children and young people online. Too many children are being exposed too early to an online world they are not ready to navigate." And in Strasbourg the day before, she went further — a direct challenge to the industry's moral ledger: "We do not have to accept addictive features. We do not have to accept children being drawn into ever more extreme content. We do not have to accept that girls have their photos used for AI-generated sexualised images."

She also gave the speech a human anchor: a tribute to Oléa, a 14-year-old Belgian girl who died by suicide about a month ago after being bullied. Von der Leyen quoted the girl's mother: "I am convinced that without these omnipresent networks, my daughter would still be here." The moment reframes this from a tech-policy file into a child-safety crusade — which is exactly how Brussels wants it debated.

Why American parents should care: the Brussels effect

The Kids Act is European law — or rather, not yet law. But American families should not assume it stays in Europe. The pattern here has a name: the Brussels effect. When the EU set the global standard on data privacy with GDPR, American platforms did not build two internets; they applied EU-grade consent banners, privacy controls and data protections to US users too, because maintaining a single global product is cheaper than a fragmented one.

The same arithmetic applies here. If Meta, Google and OpenAI must build genuine age verification, private-by-default minor profiles, stopping points in infinite scroll, and dependency-proof AI companions for the EU market, the economic case for maintaining a separate, less-safe American version is weak. EU child-safety design could become the de facto global default — including for kids in Ohio and Texas.

That said, analysis should stay honest about the limits. This is a proposal, not a law. It must be negotiated and adopted by EU lawmakers and member states — a process that, by every serious account, is years away. Platform companies will lobby, member states will water it down, and the final text could look meaningfully different from what was presented Thursday. Brussels has a long history of bold announcements that emerge from the legislative machine in a more diluted form.

The global wave is already moving

Europe is not acting in a vacuum. Australia became the first country in the world to ban social media for under-16s in December 2025. The United Kingdom will ban under-16s from social media from 2027, with default restrictions for 16- and 17-year-olds. Within the EU, France, Denmark and Greece have each introduced their own national social-media laws and pressed the Commission to act — one reason Brussels opted for an EU-wide harmonized rule rather than a patchwork of national bans.

The Commission is careful to distinguish its model from Australia's. Australia's ban gives parents no say: platforms must simply prevent under-16s from holding accounts. The EU's tiered approach — supervised mini accounts for 13- and 14-year-olds, safe-design obligations for 15- to 18-year-olds — keeps parents in the loop while putting the heaviest obligations on the platforms.

EU Kids Act FAQ: when does the under-13 ban take effect?

Not yet — and not soon. The Kids Act is a proposal, not a law. It must now be negotiated and adopted by EU lawmakers and member states, a process that by every serious account is years away, and the final text could look meaningfully different from what was presented Thursday.

What comes next: adults are next in line

The Kids Act is only the opening move. Von der Leyen said a wider framework would be announced in the autumn, aimed at adults — the Digital Fairness Act, which would take on addictive design patterns for everyone, not just minors. If the Kids Act survives its legislative journey, expect the adult-facing version to be the real fight: it would target the attention-engineering mechanics that underpin the entire ad-supported internet.

Political reaction so far splits along familiar lines. German MEP Manfred Weber, head of the European People's Party, welcomed the announcement: "Children are not an advertising target. They are our future. We should look to Australia, where users can switch off the algorithm." Others called it late — French MEP Stéphanie Yon-Courtin noted it took almost a year for the Commission to take up recommendations the Parliament had already voted on.

One thing is not in doubt: the terms of the debate have shifted. For a decade, platforms told regulators that age gates were technically infeasible and design mandates were censorship. On Thursday, the Commission formally proposed that the burden runs the other way — that Silicon Valley must prove its products are safe for children. Years of negotiation lie ahead. But the direction of travel, on both sides of the Atlantic, is now unmistakable.

Update, Sept. 17: The formal September 17 announcement added enforcement teeth the leaked draft only hinted at. Reuters reports that platforms violating the rules risk fines of up to 6% of their global annual revenue — social and video-sharing services under the Digital Services Act's penalty regime, while AI companions and chatbots fall under the AI Act's enforcement framework — and will have to pay a supervisory fee to fund regulators' oversight. Industry pushed back immediately: tech lobbying group CCIA Europe, whose members include Google and Meta, warned that mass collection of age credentials 'would undoubtedly create an attractive target for cybercriminals,' its head Daniel Friedlaender said. The Commission also confirmed AI companions and chatbots must be off by default for young users, with age checks run through certified age-assurance tools tied to the EU's Digital Identity Wallets.

Sources: Reuters (September 17, 2026); The Wall Street Journal (September 17, 2026); CNN (September 16, 2026); The Associated Press (September 16, 2026); The Times of London (September 17, 2026); Agence Europe (September 15–16, 2026). Quotes are attributed to the outlet that reported them; legislative details beyond the Commission's Thursday statement should be treated as developing. This is a developing story; TechPulse USA will update this post as the proposal moves through EU negotiations.