Social media and minors, Emmanuel Macron takes the fight for a ban to Europe

After the French measure was struck down, Emmanuel Macron has called for a European ban on social media for under-15s. Several governments support restrictions, with thresholds ranging from 14 to 16 and differing approaches to exemptions. Spain and the Netherlands also want to cover video games and AI companions that present risks. European negotiations will need to reconcile access ages, children’s rights and the product changes required of platforms, at a time when the Digital Services Act already allows action on their design.

In a letter dated 29 August and revealed by Le Monde on 8 September, Emmanuel Macron called on Ursula von der Leyen to introduce European legislation banning access to social media before the age of 15. At the same time, he confirmed his intention to redraft this autumn the national measure struck down by France’s Constitutional Council in August, with support from the Commission’s services.

The letter combines this restriction with product measures, including limits on addictive features, protections enabled by default and caps on usage time. France’s position therefore combines access controls with intervention in the mechanisms designed to prolong use.

This combination gives the debate its industrial significance. Setting an age determines who can use a platform. Regulating notifications or recommendations determines what its owner can design to retain users. Governments supporting European intervention must still agree on how to combine these two instruments.

Paris’s allies do not all support the same threshold

The French executive has set itself a deadline, aiming to complete the reform before spring 2027. The Élysée confirmed this as early as 14 August, asking the Prime Minister to draft legislation that took account of the Constitutional Council’s decision and the European framework. That political timetable now intersects with several national initiatives.

Greece supports the same threshold as France. On 8 April, Kyriakos Mitsotakis announced a ban for under-15s, targeted for implementation on 1 January 2027, while calling for a European digital majority at the same age.

Spain has set the limit at 16. Pedro Sánchez announced this direction on 3 February, asking platforms to introduce age-verification systems.

In Vienna, the governing coalition announced in March an agreement in principle for a ban before the age of 14. It is considering defining the services concerned according to the addictive characteristics of their algorithms and the content to which they expose children.

Poland’s governing coalition has also announced that it is preparing legislation targeting under-15s, with platforms responsible for age checks. Slovenia presented a similar approach in February.

Denmark shows how much the parameters can still shift. The proposal presented on 7 September sets the threshold at 15. According to analysis by law firm Plesner, the initially considered option of allowing 13 and 14-year-olds access with parental consent has been dropped. The same age threshold can therefore produce very different regimes depending on whether it includes a family exemption.

Harmonisation will require a choice between these models. It could prevent the multiplication of national controls and give the rules the reach of the European market. But it will also require every government to decide what it is prepared to change in its own proposal.

Madrid and The Hague extend negotiations to products

On 11 September, Spain and the Netherlands circulated a joint paper intended to influence the European proposal. They call for a harmonised minimum age, enforcement at EU level and an approach that could cover video games, AI companions and other services that present risks.

The document also considers exemptions for providers able to demonstrate that their service is sufficiently safe. It raises concerns about parental consent, particularly because of enforcement difficulties and additional privacy risks. The age restriction would therefore be linked to an assessment of products.

This approach would open negotiations on the evidence expected from companies. Which features should be assessed, according to which criteria and under whose oversight? An exemption based on service safety would require a process capable of tracking subsequent product updates.

In Germany, Karin Prien also supports an age limit combined with obligations imposed on providers. On 11 September, after the final report of an expert commission was submitted, the minister announced rapid measures while stressing that discussions would continue within the government and coalition. Her position signals a political direction, rather than a final federal decision.

Estonia, for its part, challenges the principle of a general exclusion. Its Minister of Justice and Digital Affairs, Liisa-Ly Pakosta, defends children’s participation in the information society and stricter regulation of platforms. Her position is a reminder that protection must also take account of minors’ rights to expression, information and social connection.

In July, Ursula von der Leyen had raised the prospect of gradual access according to age, with limited and supervised uses for younger users. This direction, supported by expert work, has yet to be translated into a proposal.

The United Kingdom, outside the European Union, offers another point of comparison. In June, the government announced a ban for under-16s, together with restrictions on livestreaming and contact with strangers, with initial measures expected in spring 2027. Messaging services such as WhatsApp and Signal would be excluded from the proposed general ban.

The French ruling requires greater precision

The Constitutional Council recognised the need to protect the best interests of the child. Its decision of 14 August nonetheless struck down a ban whose scope could have included services presenting very different levels of risk, including where those risks had not been established.

It also found that the safeguards surrounding age verification were insufficient. The obligation applied to all users, but its conditions and limits were not defined clearly enough to protect their privacy.

The redraft will therefore have to specify what is prohibited, for whom, in response to which risks and with what safeguards. The category of “social network” itself becomes a central issue. A messaging service can include public channels, a video game can host community spaces, and a video platform can combine exchange and recommendation features. Regulating according to a service’s commercial label alone would leave a rule vulnerable to product developments.

Moving to European scale does not remove this need for precision. Restrictions on fundamental rights must also comply with the principle of proportionality under EU law.

For platforms, the chosen perimeter will determine the scale of change. Closing an account, disabling particular functions or offering an age-appropriate version do not require the same systems or control procedures.

What would a new law do that the DSA does not already allow?

The European Union already has child-protection obligations under the Digital Services Act. Guidelines published by the Commission on 14 July 2025 set out the measures it considers appropriate to reduce risks including harassment, harmful content, problematic or addictive behaviour and harmful commercial practices. They support the application of the regulation. They do not amount to a general access ban based on a single age threshold.

On 10 July 2026, the Commission preliminarily found that Meta had breached the DSA because of the addictive design of Facebook and Instagram. Its findings concern infinite scroll, autoplay, notifications and highly personalised recommendation systems. According to the Commission, the company had not adequately assessed risks to users’ physical and mental wellbeing, particularly for minors, or adopted measures capable of mitigating them effectively. These findings remain preliminary.

The legislation requested by Emmanuel Macron must therefore find its place within a regulatory framework that is already being enforced. It could add a common access age, clarify the services concerned or make certain restrictions more explicit. But its relationship with the DSA will remain an open question. Which new obligations, which authority will oversee them and which criteria will determine compliance?

A uniform rule can reduce legal uncertainty, but it can also multiply administrative requirements if its obligations overlap with those already in force. For public authorities, the challenge will be to link every obligation to an identified risk. For companies, it will be to determine which product changes can genuinely address it.

Behind product features lies an attention economy

The debate over age often leaves the organisation of the user experience in the background. Infinite scroll removes stopping points. Autoplay chains videos together. A notification prompts a return to the app. A recommendation system selects content likely to generate another interaction.

These features can be useful to users, but their combination and configuration can also encourage prolonged use. It is precisely this architecture that is at the centre of Europe’s action against Meta.

In an advertising-based model, the frequency of visits and the duration of sessions can increase opportunities to display ads. An obligation limiting prompts or reintroducing interruptions can therefore alter important operating parameters. Its effects will nevertheless depend on the service, its advertising formats and user behaviour. Each additional minute does not automatically translate into additional revenue.

ANSES recommends acting first on platform design to prevent harmful uses. Its assessment calls for considering health effects in their diversity, particularly those relating to sleep, and for examining uses rather than an abstract measure of screen time.

This distinction should guide the assessment of measures. Reducing total time does not, on its own, establish whether exposure to harmful content has fallen or whether late-night use has declined. Conversely, changing a recommendation system or disabling a prompt can have an effect without reducing every use, some of which address needs for information, sociability or mutual support.

Verifying age without introducing permanent identity checks

Any age-based restriction requires a sufficiently reliable way of determining who can access a service. A declared date of birth provides information, but it does not establish that the information is true. Asking for an identity document on every platform would, however, raise questions about data collection, retention and security.

European work is intended to limit that disclosure. The solution presented by the Commission allows a user to prove that they are over 18 without sharing other personal information with the service concerned. Initially designed for adult-only content and services, it is intended to be compatible with future European digital identity wallets. The Commission states that it can be adapted by Member States and market participants.

Meta has already negotiated product changes in the United States

The US case has moved a step further. On 26 August, Meta announced an agreement intended to settle allegations by US states concerning the effects of Facebook and Instagram on younger users. It provides for changes to the services, including daily usage limits and overnight restrictions. The company continues to deny wrongdoing.

The agreement is therefore not a judgment establishing all the plaintiffs’ allegations. It nevertheless shows that judicial pressure can lead to commitments that directly affect the user experience.

For Europe, the comparison lies in the effective transformation of products. Whether an obligation results from a settlement, an enforcement decision or new legislation, its effectiveness will depend on its scope, duration and the ability to verify its effects. A commitment on usage time and a measure on the functioning of recommendations address different problems. Their implementation will need to be assessed separately.

Effectiveness will be measured in actual use

A common European threshold would offer an immediately understandable rule. Its effectiveness would nevertheless depend on the chosen scope, exemptions, the reliability of controls and the services towards which minors might move. Graduated protection would, for its part, require criteria to determine what becomes accessible at each age.

Design obligations raise another measurement issue. Should authorities observe late-night use, repeated prompts, exposure to certain content or the ease with which a session can be interrupted? No single indicator will be sufficient to describe a service’s safety.

A policy combining access restrictions and product transformation will therefore need to provide for its evaluation from the outset. Platforms should be able to document the effects of the changes requested. Authorities and researchers should have access to the information needed to assess them. Families will need understandable settings and usable routes for redress.

Ursula von der Leyen’s State of the Union address, scheduled for Wednesday 16 September, should clarify the direction she intends to take. National differences give a sense of the trade-offs that will follow.

The operational translation will then be decided through far more concrete choices. A notification sent or suspended. A video launched automatically or left to the user’s initiative. Proof of age limited to the information that is necessary. This is the level at which the promised protection will begin to become observable.