EU child social media ban recommended

The recommendation for “a harmonised EU-wide access restriction to social media and other digital services, including AI companions, for children under 13” was set out in a new report published by the co-chairs of an expert panel set up by the European Commission last year to advise it on child online safety. They published their final report on Monday.

Dr. Maria Melchior, director at the French National Institute of Health and Medical Research, and professor Dr. Jörg M. Fegert, director of the Department of Child and Adolescent Psychiatry, Psychosomatics and Psychotherapy at Ulm University Medical Centre, said children under the age of 13 in the EU should only have access to social media and other digital services in limited circumstances. Those include where the digital environment is “age-appropriate”, access is “time-limited”, and if there is either “parental authorisation and supervision” or access is enabled “in educational contexts”.

Melchior and Fegert said the restriction should be underpinned by “proportionate” but “effective” age-assurance systems.

For children aged between 13 and 18, a further set of measures have been recommended. Melchior and Fegert advocated “limits on infinite scrolling, autoplay, push notifications, and problematic personalised recommendation engines” and said “protective default settings” should also be set to “tackle addictive design, dark patterns, unsolicited contact and rabbit holes”. These additional measures should “adaptable”, they added, so as “to continuously address new harmful design features”.

Under Melchior and Fegert’s proposals, providers of digital services would have to prove “their products and services are safe for minors” before they could make their services accessible to children.

“Specific age restrictions should apply until key safety design features protecting children and adolescents are introduced by social media and other digital services providers,” the co-chairs said. “Until providers have demonstrated safe and age-appropriate features, therefore complying with EU and national measures, they should not have access to minors. Where there is a ‘safe by default’ version, age-inappropriate features (features for adults) should only be activated following effective age assurances.”

Individual EU member states should be given scope to apply additional precautionary access restrictions to social media and other digital services for adolescents aged 13 or over, they said.

In their report, Melchior and Fegert said the risks of children encountering criminal and illegal content online are “considerable” and that they also face “extensive” risks to their physical and mental health from using digital services. However, they also acknowledged that children can benefit from online access, if services are ‘safe by design’.

“Social media and other digital services that are ‘safe by design’ could contribute to children’s and adolescents’ empowerment and participation online by supporting the acquisition of digital skills and media literacy,” the co-chairs said. “Age-appropriate digital education delivered by schools with the support of parents and caregivers could also equip minors to seize the opportunities of the online world and to manage possible risks and harms.”

Wesley Horion of Pinsent Masons said: “The report of the expert panel points towards a more interventionist and preventive compliance model. Platforms may increasingly be expected, not merely to mitigate risks once identified, but to demonstrate that services and individual features are safe and age-appropriate before making them available to minors. This would make child safety a product-design and market-access issue, with consequences for recommender systems, engagement features, default settings, and age assurance.”

“However, allowing member states to impose additional restrictions above the proposed EU-wide threshold of 13 could undermine that model: platforms might face different age limits, exemptions and verification requirements across 27 jurisdictions. Such fragmentation would increase compliance costs without necessarily improving protection. The Commission should therefore resist carrying this aspect of the report wholesale into the Digital Fairness Act, a consumer protection-focussed proposal expected by the end of the year, and instead pursue a harmonised, risk-based framework that gives industry a clear and consistent standard to meet,” he said.

From a child-protection perspective, however, the co-chairs are right to reject a binary choice between unrestricted access and blanket exclusion, according to Horion.

“The report’s developmental approach recognises that safeguards should evolve as children acquire greater autonomy, while placing primary responsibility on the businesses that design and control digital environments,” Horion said. “Yet safer products and age checks alone will not equip children to recognise manipulation, think critically or pursue a responsible online behaviour. Effective protection will also depend on parents, teachers and educators being able to support that transition, despite many lacking the necessary training or confidence themselves.”

“The challenge for policymakers is therefore to combine enforceable and harmonised duties for industry with sustained investment in digital and AI literacy and adult capacity-building, rather than allowing either side to become a substitute for the other,” he added.

On Tuesday, a committee of MEPs also called on the European Commission to intervene on child online safety, including by introducing personal liability for directors of social media platforms. The European Parliament as a whole is expected to consider the report in September.

Late last year, Australia became the first country in the world to ban children under the age of 16 accessing social media platforms. Policymakers in other parts of the world have been considering similar action.

In June, the UK government outlined its own plans to prohibit social media platforms from offering their services to children under the age of 16, from spring 2027. Additional measures, including blocks livestreaming and stranger communication with children for under-16s, minimum age requirements of 18 for AI chatbots that either operate as a “romantic companion” or have “similar intimate functionalities”, and overnight curfews and breaks in infinite scrolling for under-18s, are also under consideration. The UK government is expected to set out more details on those measures this month.