Europe Proposes a New Test for Big Tech: Prove Children Are Safe by Design

Europe wants child safety built into digital platforms and the biggest providers to prove it. The proposed EU KIDS Act would combine a minimum age of 15 for independent social media accounts with curbs on addictive features and AI companions that create emotional dependency, putting platform design at the heart of online child protection.

The European Commission has proposed a child-safety framework that would change both when children can enter social media independently and what platforms can offer them once they arrive. The EU KIDS Act would prohibit social media access for under-13s, introduce supervised accounts for younger teenagers and impose design obligations across services used by minors. The proposal now goes to the European Parliament and Council for examination and adoption.

Families would receive more controls, while providers of very large online platforms would have to demonstrate that their services are safe by design. "It is for platforms to show they are safe by design," Commission President Ursula von der Leyen said. The proposal would make the design of digital services a direct subject of compliance scrutiny, alongside children's eligibility to use them.

Age limits redraw the route into social media

Under the proposal, children could open independent social media accounts from 15. Those aged 13 and 14 could access age-appropriate social media and video-sharing services through "mini accounts" created by guardians and accessed through the guardian's account. These services would have to restrict social contacts and limit screen time to one hour a day, placing supervision within the account structure itself.

Children aged three to 12 would have a narrower route: specially designed, child-friendly video-sharing services accessed through guardian-managed accounts. Platforms would need to provide an easy-to-use tool that restricts an adult's device to those services when it is handed to a child, with exposure capped at one hour daily. Social media would remain unavailable to this age group.

The proposal would recognise different levels of independence while preserving controlled access to selected services. Reaching 15 would permit an independent social media account, but the broader protections for under-18s would continue. Eligibility to register and the obligations governing a young person's experience would therefore remain separate regulatory questions.

For parents, the practical test would be whether these controls can be used consistently without requiring detailed technical knowledge. Limits would also need to operate meaningfully when children move between accounts, devices and services. The effectiveness of this supervised model would depend on the relationship between household oversight and platform design: parents would manage access within environments that providers must make suitable for children.

The feed itself comes under scrutiny

The safety obligations would extend across social media, video-sharing services, online video games, AI companions and chatbots offered to under-18s. They would prohibit addictive features and profiling-based recommendation feeds that draw minors into harmful content. Infinite scrolling without stopping points, reward tricks, notifications during sleeping hours and unsolicited contact from strangers would also fall within the proposed restrictions.

These provisions target the mechanisms that organise attention and interaction. A service could meet an age threshold yet still expose young users to features covered by the design rules. Compliance could therefore require changes to recommendation systems, notification schedules and interaction settings, in addition to registration checks. For covered businesses, child safety would become a product-development obligation with implications across several parts of a service.

Minors' profiles would be private by default, with geolocation, camera and microphone access switched off. Providers would also have to offer effective time-management tools, simple ways to block or mute users, and safer recommendation systems that minors can control, adjust and reset. Default settings are central to this approach because they determine the starting conditions of participation before a child or guardian changes anything.

AI companions and chatbots would be switched off by default and barred from simulating interpersonal relationships in ways that create emotional dependency. This brings the nature of an interaction, as well as the content it produces, into the safety assessment. Enforcing such a restriction would require workable criteria for judging when conversational design encourages dependency, making the interpretation of the obligation consequential for providers and auditors alike.

Age checks must earn users' trust

The access rules would depend on services being able to establish who is a child. Online services and app stores would have to use age-assurance tools, while social media and video-sharing providers would verify age when users create accounts. For existing accounts, the proposal envisages estimates based on reasonable indicators, including account creation dates and credit card details.

The Commission identifies the EU age-verification app as one option and says it does not retain identity documents or biometric data. Member States would help establish the wider age-assurance system. The intended combination is consequential: providers would need enough confidence about age to apply restrictions while limiting the personal information involved in reaching that judgment.

Reliability and privacy would have to be assessed together. Incorrect age estimates could deny legitimate access or fail to apply protections to children, while shared devices and accounts would complicate the relationship between account information and the person using a service. The handling of errors, challenges and repeated checks would influence whether age assurance becomes a credible safeguard that families can use with confidence.

Safety claims face an enforcement test

The proposal would require providers of very large online platforms to submit compliance plans to the Commission and an independent auditor, with new services, features or functions subject to assessment. The Commission could request proposed corrective measures where an auditor's report identifies shortcomings. This creates a specific oversight mechanism through which claims of safe design would face scrutiny beyond the provider's own assessment.

Enforcement would build on structures established under the Digital Services Act and Artificial Intelligence Act. The proposal envisages expedited investigations into noncompliance, with the Commission expected to conclude them within 90 days. Speed could help address shortcomings sooner, but the quality of audits, clarity of standards and effectiveness of corrective measures would remain essential to the framework's practical value.

The Commission cites the Special Eurobarometer on the Digital Decade 2026, according to which 92% of Europeans regard stronger online protection for children and young people as a top policy priority. The figure indicates broad demand for protection, without establishing agreement on every proposed instrument. Age thresholds, parental responsibilities, privacy safeguards and restrictions on product features each raise choices that the legislative process must examine.

Parliament and Council will now determine how those choices take legal form. The crucial test is whether the eventual rules connect age-based access, usable parental controls and enforceable design standards into a coherent system. An age check can determine who enters a platform; sustained protection will depend on what the platform permits, encourages and prevents after that entry.