European Commission President Ursula von der Leyen recently laid out a sweeping vision for regulating children’s access to social media. Drawing on a report from a special EU panel on child safety online, she called for a “phased and gradual access” model: no screens at all for children under three, time-limited and supervised access for kids under 13, and gradual, “age-appropriate” access for teenagers.
She also argued that tech companies—not parents—bear the primary responsibility for making platforms safe, comparing it to car manufacturers being required to install seat belts and airbags rather than leaving safety up to drivers.
Let’s be clear about one thing first: Protecting children’s developing minds matters enormously. Research on social media’s effects on adolescent mental health, attention, and self-image is alarming, and no reasonable person wants children exposed to predatory algorithms, exploitative design, or harmful content. Any fruitful policy conversation about this issue must start from that shared concern.
But shared concern about a problem doesn’t justify policies like this EU push, which raises serious red flags.
A troubling precedent against free speech and access to information
Social media, despite its flaws, is also a platform for expression, association, and access to information. Government-mandated age gates enforced at the platform level require identity-verification infrastructure that inevitably touches everyone, not just minors.
Once the EU builds the legal and technical machinery to decide who may speak or listen online and at what age, that machinery doesn’t disappear—it becomes a permanent feature of the internet that can be tightened, expanded, or repurposed by future officials with different priorities. A policy justified today as protecting toddlers can easily become the foundation for broader speech control tomorrow.
Shifting responsibility away from parents
Von der Leyen’s seatbelt analogy is appealing, but it doesn’t quite hold up. Cars are dangerous in a narrow, mechanical way—a manufacturer either builds a safe frame or doesn’t. Childhood development is different: It depends heavily on individual family values, maturity, and context. Parents—not Brussels—are best positioned to know whether their 12-year-old is ready for a limited Instagram account or whether their 15-year-old needs firmer boundaries.
Centralizing that judgment in a continent-wide regulatory scheme substitutes one-size-fits-all bureaucracy for the nuanced, ongoing job parents already do every day. It risks infantilizing families and signaling that the state, not the household, is the primary authority over how children are raised.
Blunt instrument for a nuanced problem
Blanket age thresholds don’t distinguish between a teenager using social media to stay connected with a school club and a teenager doomscrolling harmful content for six hours a night. The actual harms—addictive design, exploitative algorithms, inadequate content moderation—are platform behaviors that could be targeted directly through transparency requirements, design standards, and enforcement against demonstrably harmful practices, without needing to restrict access itself or build age-verification systems that inevitably raise privacy concerns for adults too.
Likely won’t work as intended
Countries that already implemented age bans, including Australia, experience predictable problems: teens circumventing verification with VPNs or fake credentials, verification systems creating new privacy and data-security risks, and enforcement burdens falling on platforms in ways that are inconsistent and difficult to audit.
None of this means the status quo is fine. Tech companies could do far more to protect young users, and parents deserve better tools—such as default privacy settings, transparent algorithms, and easier reporting—to help them make decisions for their own families.
But the answer to a real problem shouldn’t be a continent-wide access-control regime that treats free expression as fungible and hands the state a permanent role in policing who gets to participate online. Better defaults, real transparency, and support for parental choice and education would do more good with far less risk to the open internet that both children and adults depend on.