Whether it’s age restrictions, offline modes during nighttime, or special platform settings: No current issue in digital policy is debated as intensely as the influence of social media on the lives of children and young people. Parents have reason to be concerned, while politicians continue to promise that they will finally act on the matter. Yet the youth that is at the center of this discussion are often only consulted for the support of certain narratives – and not involved in developing solutions.

In June 2026, the German expert commission “Protection of Children and Youth in the Digital World” (“Kinder- und Jugendschutz in der Digitalen Welt“) submitted a list of 56 recommendations to Karin Prien, the German Minister for Family Affairs, based on an assessment published in April of the same year. Perspectives of minors were incorporated in the process. On September 11, the experts presented their final recommendations to the public. The minister is now tasked with deciding which of the recommendations she will implement. During the presentation, she reiterated her preference concerning the question of age restrictions: In Germany, children and adolescents shall be gradually restricted from having access to social media platforms. However, these restrictions are not meant to be the only policy solution but could be tied to new obligations regarding the appropriate age design of platforms.

In political terms, the recommendations form the basis for a more general strategy of the federal government. Minister Prien also announced plans to push through a temporary regulation on the national level, for which a paper on the key points is already being discussed, while awaiting an initiative from Brussels for a common European approach – which followed suit: By announcing the EU KIDS Act, the EU plans to limit social media access to teenagers in the Union from the age of 13, with personal accounts being allowed only at age 15.

In our view, blanket age bans and limits do not solve a single concrete problem we have with platforms; they are a mere tokenism. Age limits exclude young people, while platforms are left to decide whether to improve their services to make them age-appropriate. While regulations to correct platform behavior already exist, they must also be consistently enforced, and platforms in turn must be compelled to comply. For this reason, it is worth taking a closer look at the recommendations by the German expert commission, comparing them with the EU KIDS Act that was recently announced as well as the European Union’s existing guidelines, and finally assessing which requirements are most likely to ensure a lasting and democratically aligned protection of children and young people on the internet.

Children and young people are not the problem. It’s the platforms.

Focusing solely on screen time and limited social media use for young people misses the heart of the debate. A significant portion of the risks posed by social media does not stem from individual user behavior but is part of the product and market strategy of major tech platforms: Endless feeds with no stop mechanisms, autoplay features that start the next recommended video seamlessly, push notifications that urge users to return, and algorithms optimized for attention rather than well-being.

As of late August 2026, the U.S. is providing a glimpse into what a potential “fix” for the system behind social media platforms might look like. To end a lawsuit involving a total of 29 states, Meta agreed to a settlement involving payments of up to $18 billion as well as changes to its services. The company behind Facebook, Instagram, and WhatsApp was accused of deliberately using attention-maximizing designs on its platforms that encourage harmful behavior and can lead to mental health issues such as depression, anxiety disorders, and suicidal thoughts, thereby putting young people at risk. The required changes on Instagram and Facebook are to be set as the mandatory default for young people and monitored by an independent oversight body. The settlement includes a series of measures for increased youth protection: a daily limit of two hours for 13- to 17-year-olds, a nighttime restriction from midnight to 6 a.m., a school mode with reduced notifications during school hours, hidden “like” counts, disabled beauty filters, autoplay features that can be turned off, and the option to select an alternative feed without algorithmic recommendations. It is worth noting that these changes were not enforced through the threat of an age restriction but rather through pressure of liability. This suggests that the product design can indeed be regulated – if the will exists.

However, the settlement does not set a model precedent, as researchers and legal experts explain. In fact, the terms were co-negotiated by Meta and not mandated by lawmakers. At the same time, Meta has secured a payment suspension for parts of the settlement until its competitors follow suit and implement the same measures. This way, Meta could use its own protection model as leverage against the industry, even though it has not yet been verified that the planned mechanisms for youth protection are at all sufficient. Furthermore, the changes are based on Meta’s own age estimates of its users: According to the Electronic Frontier Foundation, the company calculates an estimated age by collecting even more data for analysis – potentially including the monitoring of adult accounts to verify whether they might in fact belong to minors.

Platforms know of their wrongdoings. They do it anyway.

In the U.S., only legal proceedings could achieve what’s already been provided by dedicated rules in the EU. For instance, Article 28 on the online protection of minors in the Digital Services Act (DSA), the European legislation governing online platforms. Still, enforcement is the main issue here. All the while, a field study conducted by the German Federation of Consumer Organizations (vzbv) shows that social media platforms do not adhere to the rules even with knowledge of their users’ age. During the test, five major online platforms were examined using dummy accounts that corresponded to the profiles of 13-year-olds. Despite the DSA regulations and accompanying guidelines, compulsive design mechanisms were found on all five platforms, including “infinite scroll” (a never-ending feed of content) and the autoplay feature. Several platforms asked for a reason when the restricted mode – which was turned off by default – was activated. Predefined time limits could be bypassed for the respective day of use through a specific feature. None of the platforms provided a non-personalized feed as the default, and not all of them allowed users to manually switch to such a feed. Where this non-personalized feed could be activated, pop-ups or a blank homepage subsequently appeared.

The tricky part: Mandatory online identification

Concerning the introduction of an age limit or banning access to social media platforms for certain age groups, the German experts were undecided. As a result, two possible alternatives are on the table: The first suggestion proposes a minimum age of 13 for using online platforms, with gradually changing protection standards up until the age of 18 (the age of legal majority in Germany). The second one argues to forego a uniform age limit and instead recommends tightening the DSA regulations to restrict user accounts, services, or individual features where they pose particular risks. We at AlgorithmWatch consider the second suggestion to be the more viable one, as it follows an approach oriented towards the risks and the platform design rather than treating online services – which vary greatly from one another – the same way across the board. If platforms fail to comply with these obligations, they must face stricter sanctions. Minister Prien, on the other hand, has already spoken out in favor of the first option, both publicly and in discussions with the European Union.

In this context, it is worth looking at Australia: Since December 2025, the country has set a minimum age of 16 for certain platforms. Three months after the new law took effect, the Australian regulatory authority’s initial evaluation shows that the proportion of 10- to 15-year-olds who have their own account on age-restricted platforms only fell by roughly 10 percent, while 81.5 percent continue to use the platforms regardless of whether they have their own account or not – primarily because they have not yet been asked to verify their age. Here, too, inadequate implementation by the platforms seems to be the problem. At the same time, the report documents several side effects: The percentage of parents who were unaware of their children’s social media use rose by ten percent, while the overall use of messaging services and online games increased – possible indicators of decreasing trust in the parent-child relationship. On matters of well-being or mental health, the evaluation found no changes. According to the report, the law has had no impact in these areas so far.

Meanwhile, both options proposed by the German expert commission depend on the same condition: an effective age verification. It thus becomes a de facto prerequisite for structural safeguards to take effect. We cannot fully endorse this approach: From AlgorithmWatch’s perspective, age verification should merely be a “last resort” option. In an ideal world, in which platforms fulfill their design obligations to mitigate risks, age verification would not be necessary at all. However, as long as this does not happen, we see it as our responsibility to ensure that emerging systems ensure the utmost protection of the privacy and interests of their users – especially where they are considered to have no alternatives. Given that, to verify whether users of an online service are under 18, all users must be validated. Age verification for minors is therefore equivalent to a mandatory ID requirement for adults – a fact we should all be aware of.

The basic framework for how this monitoring might work in practice has already been established: Both the EU and the German minister for family affairs are relying on the EUDI Wallet, a government app for digital IDs that is scheduled to be available in Germany starting in early 2027. However, a parliamentary question submitted by the Green Party in late July 2026 (Parliamentary Document 21/7139) revealed that the EUDI Wallet is still not fully developed, both technologically and in terms of data protection. The EUDI Wallet has faced significant criticism as a tool for age verification, precisely because the conditions under which age verification would be acceptable have not yet been met. According to the expert commission, the minimum requirements for age verification that protects personal privacy include data processing on the user’s own device, data minimization, selective disclosure of data, and the exclusion of biometric and behavior-based methods. These requirements must be ensured before age verification can become a prerequisite for access to online platforms.

Having platform providers conduct age verification instead is no solution either. Currently, platforms rely on behavior-based use analyses to estimate age. Anyone who now shares additional reliable data on their age with these non-transparent systems could enable even more precise targeting and profiling – even if this is prohibited by law. Examples from Australia and the U.S. show that platforms do not always adhere to the guidelines they are supposed to follow.

What the expert commission is doing right

In number 37 of the 56 recommendations (in German “Handlungsempfehlung”, HE for short), the expert commission by Federal Minister Prien presents a binding catalog of safe default settings for social media platforms – a kind of protection of minors “by default”. On social media accounts used by children and young people, personalized advertising should be disabled as a standard, and “dark patterns” (design patterns that deliberately manipulate users into taking specific actions) and features that promote compulsive behavior should be prohibited. This is an important point that addresses the structural level rather than placing the responsibility solely on individual media literacy. These settings would ensure a more self-determined use of social media, regardless of the users’ age, empowering consumers to make their own decisions rather than having them dictated by platform mechanisms. Recommendation 39 has a similar goal regarding safe default settings for algorithmic recommendation systems, mandating understandable and age-appropriate settings for users – a long-overdue clarification of the DSA rules on recommendation systems (Articles 27 and 38).

Instead of merely setting limits, the experts also emphasize support: age-appropriate services, open European infrastructures, independent long-term research – including access to research data as per Article 40 of the DSA – and the mandatory involvement of young people in the entire process are also among the recommendations (HE 47, 50, 52–54). In addition, the experts warn against the precautionary blocking of permissible content (HE 51), as the protection of minors in the media must not become a tool for censorship or the restriction of democratic discourse.

Social media: A case for pediatricians?

Another point of criticism concerns the framing of the expert commission’s recommendations. In particular, the recommendations for 10- to 12-year-olds prioritize the prevention of excessive social media use (HE 27) and closely link media education with health (HE 28) – going so far as to suggest involving pediatricians as trusted advisors for the protection of children and young people from harmful media content. In their own assessment, however, the experts note that it is not the duration of use but rather the quality and content that determine the impact of social media. Targeted services for those affected as well as at-risk groups are therefore important and sensible. However, medicalizing platform-related consequences shifts responsibility back to the individual. The correct approach would be to disable compulsive features by default, rather than declaring the use itself a clinical condition.

At the EU level: Legislation with age restrictions

The debate in Germany is not a purely national one: the EU and its member states are also discussing how children and young people can be better protected on social media. In her State of the Union address on September 16, 2026, von der Leyen announced the “EU KIDS Act”. At its core, the proposal aims to limit access to social media: Ban access for children under 13, “mini accounts” for those aged 13 to under 15 set up and supervised by parents, individual accounts only for those aged 15 and older, and a requirement for platforms to offer safe design.

The proposed legislation stipulates that children under 13 may only use child-appropriate services through their legal guardian’s account– meanwhile such services have neither been designated nor developed yet. The “mini accounts” are not a mandatory requirement but rather an exception that platform providers may choose to implement. The parental responsibility on which the regulation is based therefore only applies to the extent that a company has previously permitted it. Any company wishing to take this route must submit an impact assessment and implement additional design requirements targeting recommendation systems and personalization, among other things.

However, simply implementing these requirements cannot automatically be considered an effective risk mitigation measure. This framework could provide an incentive for platforms that wish to continue reaching the relevant target audience to make age-appropriate improvements to their services. Another possible consequence, however, is that providers simply choose not to offer these options. The result would then resemble a ban without formally being called one, since it ultimately amounts to a product-based decision rather than a regulatory one. Furthermore, the draft legislation lumps together various services – including social media, online games, and AI applications – which are themselves subject to other, distinct laws (such as the AI Act). The EU Commission demonstrates in its own planning that there is another way: The announced “Digital Fairness Act”, which is intended to regulate manipulative design practices, influencer marketing, and other aspects for all consumers – without any age restrictions whatsoever.

This new regulation is partially based on the work of an EU panel. Its report recommended an EU-wide harmonized age restriction starting at age 13. At the same time, the experts also called for a reversal of the burden of proof, whereby it is not up to regulatory authorities, parents, or children to prove potential harm. Rather, providers must guarantee the safety of their services before they are even allowed to reach minors. Similar to the German expert commission, the EU report also calls for binding rules for the design features and platforms’ recommender algorithms. At the same time, it identifies attention-maximizing business models, harmful design features, lack of compliance, and insufficient access to research data as problematic areas that cannot be resolved by blanket age limits. The report also acknowledges that young people usually know how to circumvent blanket age limits. The EU KIDS Act aims to counter such circumvention with corresponding EU-certified age verification systems – that we are so far waiting to see.

The process and the timeline for implementing new regulations within the EU are also being criticized: Following individual actions by single member states (including France, Denmark, and Greece), technical solutions for age verification were presented on short notice, while tech-savvy nations and experts consulted by the panel clearly advocated for a more nuanced approach. This creates the impression that swift political action and a personal political agenda were more important than a sustainable and technologically secure improvement of the online realities for users.

Right direction, wrong route.

The assumption that verifying a user’s age is enough to enable safe media use is oversimplified, because implementing a sustainable improvement is significantly more complex. When it comes to digital services, however, we usually don’t discuss consumer protection until the product is already on the market. This would be unimaginable if we were instead talking about toys that are meant to be sold. Teenagers do not have to adapt to the logic of platforms. Instead, social media providers must meet the needs of children and young people. As long as platforms are not urged in a determined and timely manner to fulfill their obligations and ensure safe settings, age verification or separate youth accounts will be of little use. The guiding principle must be platform responsibility instead of blanket bans or even the exclusion of entire user groups.

At the EU level, the right tool is already available. Article 28(1) of the DSA requires all online platforms accessible to minors to take appropriate and proportionate measures to ensure a high level of privacy, security, and protection. The European Commission has explained in its guidelines what this means: age-appropriate default settings such as a private account, no autoplay, no push notifications or contact recommendations, no tracking, no harmful design, effective reporting channels, and age-appropriate recommendation systems based exclusively on stated preferences rather than inferences drawn from user behavior. In addition, Article 25 of the DSA prohibits the use of dark patterns.

When it comes to age verification, the EU guidelines are also significantly more cautious than the political debate: they are considered proportionate only where clear legal age limits already apply – for example, for pornography or games – or where specific risks cannot be mitigated in any other way. They do not serve as a substitute for secure platform design, as age verification alone is not sufficient to comply with Article 28. Furthermore, age verification systems must meet certain criteria: Among other things, providers may not use age verification to gather additional data for identification, geolocation, behavioral profiling, or tracking. Here, the legislature has already recognized that youth protection can be used to collect data and is taking steps to prevent this.

Hence, much of what both the EU panel and the German expert commission call youth protection "by default" is already articulated in these guidelines. What is lacking, however, is their concrete enforcement. While various experts argue that the Commission’s guidelines do not constitute legally binding rules, they clearly indicate the criteria the European Commission uses when evaluating platform violations. The bottom line is: Both the German expert commission and the EU panel provide a well-founded, structurally sound framework for addressing a complex problem. However, contradictions arise when the experts water down their risk- and design-oriented approach with blanket age limits and mandatory age verification – and when politicians only select those alternatives from experts that promise them quick wins. A lack of knowledge about users’ ages is not the problem why platforms remain unsafe, not just for younger audiences, but for all audiences. It is rather the lack of willingness from platform providers to act accordingly to what they already know. The decisive lever, therefore, remains platform accountability: Providers are the ones who need to be sanctioned for their failure to comply, not children and young people who did nothing wrong.