On 25th March 2026, a Los Angeles court found social media giants Meta and YouTube to be liable for deliberately designing addictive platform features that harm users, particularly minors. Some commentators compared this to the landmark litigation against the tobacco industry in the late 1990s, in which tobacco companies were held accountable for the addictive and harmful nature of their products. The tobacco case marked the beginning of a series of regulatory and preventive measures that, over decades, significantly reduced tobacco consumption and thus mitigated its harmful effects. Some observers now anticipate a similar trajectory with respect to social media, while others caution against overstating the comparison, given the complexity of social media’s effects on consumers.
The Los Angeles ruling reflects a broader global trend towards greater regulation of digital platforms. In late 2025, Australia became the first country worldwide to introduce a statutory minimum age for social media. Users under the age of sixteen are prohibited from holding accounts on major platforms, and enforcement responsibility lies with the companies themselves, which face substantial financial penalties for non-compliance. Across Europe, similar proposals are gaining momentum, with discussions ranging from stricter age limits to enhanced parental consent requirements. The regulation of minors’ access to social media has thus moved from a policy concern to a concrete legislative agenda.
The debate’s particular focus on the protection of minors stems from the widening crisis in youth mental health, as highlighted by the Lancet Psychiatry Commission on Youth Mental Health. According to this commission, research suggests an ongoing decline in mental health among adolescents over the past two decades, entering a dangerous phase. The causes appear multifactorial. Besides major global crises such as the COVID-19 pandemic and climate change, increasing attention has been directed toward the largely unregulated rise of social media as a potentially contributing factor.
But why exactly are adolescents so vulnerable with respect to their developmental psychological characteristics, and what challenges does this vulnerability pose for regulation and policy?
Psychological Analysis
Developmental Psychology of Adolescence
Adolescence is a distinct stage of life, defined by unique and sometimes puzzling psychological characteristics, such as heightened emotional reactivity and self-consciousness. During this period, emotional responses can be especially intense and impulsive, while self-regulatory capacities are still maturing. Increased self-consciousness heightens concern with how one is perceived by others, and peer approval and rejection can therefore carry particular significance. A casual remark on one’s appearance, a perceived negative look, or exclusion from a social activity can carry disproportionate emotional weight during adolescence. Consequently, social comparison tends to be experienced more intensely than in other life stages.
Adolescence is also a time of identity exploration, as one’s self-concept is still developing and social feedback therefore plays a stronger role in shaping the self-perception. For example, experimenting with a new interest, style, or opinion is often part of testing different social roles before committing to a more stable sense of self. For an overview of adolescent developmental psychology, see Steinberg (2020).
Social Media’s Architectural Design
The architectural design of social media platforms can influence how adolescents experience these developmental sensitivities. Social media platforms are structured around engagement optimisation, meaning that their algorithms are designed to prioritise content that sustains attention and interaction. The more time spent and the higher the user engagement, the better the platform’s economic performance.
The way platforms optimise engagement bears an interesting resemblance to a well-established psychological principle called intermittent reinforcement, whereby rewards are delivered unpredictably rather than consistently. Unpredictable rewards are associated with activation of dopaminergic reward pathways, i.e. neural systems involved in motivation and learning, which current evidence suggests are particularly responsive during adolescence. Think of scrolling through a social media feed: the next post might be a boring video that does little to your experience – or it can be unexpectedly captivating. It is precisely this unpredictability that sustains repeated checking behaviour. During adolescence, these reward systems are especially responsive, which may intensify the effects of such design choices, thereby keeping adolescent users on the platform longer. Continuous availability, often embodied in “infinite scroll” interfaces, further feeds into this engagement optimisation logic. Such online environments lack natural stopping cues, making disengagement more difficult, especially during a developmental stage when regulatory capacities are still maturing.
Another central feature of social media platforms is that social approval is both quantifiable and public. While in “real life”, signs of social approval are often more ambiguous and context-dependent, they become quantifiable through metrics such as likes and follower counts. Upward comparison – comparing oneself with others perceived as superior in, e.g. popularity – can be amplified in such environments, as users are continuously exposed to high-performing content. Thus, a gut feeling of being less popular than others can suddenly become an apparent, quantifiable truth, particularly during adolescence, when social comparison carries heightened emotional weight. Whether these metrics actually capture social standing accurately is secondary to their perceived significance. In addition, feedback becomes public and persistent. Offline interactions often remain in small social circles and fade over time. Online, they can reach wider audiences and may remain accessible indefinitely. Further, due to social media’s public nature, there is a collapse of social contexts. Offline, individuals navigate distinct social spheres, such as family, peers, school, and strangers, each defined by different social norms. Online, these boundaries blur, and audiences often collapse into one. At a stage marked by identity exploration, like adolescence, such blurred boundaries may complicate the development of distinct social roles.
Social Media and Well-Being
When these design choices intersect with adolescents’ developmental sensitivities, their effects may be amplified. Engagement-optimised design aligns with heightened reward responsiveness and still-maturing self-regulation. Publicly quantifiable feedback may intensify peer approval concerns, while the collapse of distinct social contexts can complicate identity exploration and formation. In other words, when developmental sensitivities meet social media’s design, this may have consequences for adolescents’ well-being. But how far does the available evidence support these theoretical concerns?
Given the relatively recent rise of social media as a central pillar of our daily lives, a review of the available literature across the behavioural sciences on social media use and adolescents’ well-being and mental health paints a nuanced yet partly limited picture. Sala et al. (2024) provide a comprehensive overview of the current evidence base on social media use and adolescent well-being. On the one hand, moderate use of social media (approx. under two hours per day) appears to be associated with benefits. Platforms can create a sense of belonging, offer opportunities for identity exploration, and provide access to peer support networks, especially for marginalised groups, while also facilitating access to information, especially in contexts where media freedom is limited.
Despite these benefits, however, several concerning trends can be outlined. Younger adolescents (around 10 to 15 years old) appear to be especially vulnerable to experiencing negative consequences with regard to their well-being, including increased depressive and anxious symptoms. This vulnerability is often attributed to less mature social and mood regulation skills in this age cohort, as elaborated above. While moderate use of social media may be associated with certain advantages, heavy use (approx. two or more hours per day) appears to be linked to adverse developments, including lower well-being, happiness, and self-esteem, as well as poorer sleep and more negative body image. Some evidence also suggests that the risk of depression increases with the amount of time spent on these platforms.
One potential explanation for this may be the displacement hypothesis, whereby social media use replaces health-promoting activities, such as in-person interaction and physical activity. Further, the relationship could be bidirectional: adolescents experiencing depressive symptoms may engage more with social media, which can further increase such symptoms. Anxiety shows similar patterns, particularly when platform feedback mechanisms heighten fears of social rejection. While concerns about social media addiction are growing, it is currently not yet recognised as a formal disorder in the DSM-5-TR (the standard mental disorder classification used by clinicians).
Despite evidence being somewhat limited – partly due to the relative novelty of social media, the difficulty of establishing causation, and limited access to the platforms’ data – the potential consequences are serious enough to take action. The Leopoldina Foundation, in its 2025 report on the online protection of minors, argues that the available evidence, combined with potential irreversibility of psychological harm during critical developmental periods, justifies the application of the precautionary principle. This principle is a well-established legal and ethical standard and frequently cited by EU policy. It claims that when suspicion of significant harm is scientifically grounded, responsible actors can and often should take protective measures, even when the body of evidence is not yet conclusive.
Policy Implications
Translating psychological insights into policy requires distinguishing between different types of intervention. If the risks associated with adolescent social media use arise partly from platform design, regulatory tools targeting market actors become relevant. If, by contrast, the concern lies in developmental sensitivity and patterns of use, measures related to education and guidance come into focus. Within the European context, this distinction is institutionally significant. While the regulation of digital markets falls largely within EU competence, education measures remain primarily the responsibility of Member States. The level at which action is taken, therefore, shapes the type of intervention.
Platform Interventions
At the structural level, the EU has already established a regulatory framework through the Digital Markets Act (DMA) and the Digital Services Act (DSA). While the DMA primarily addresses market contestability and gatekeeper power, the DSA is more directly relevant to the protection of minors. It introduces a risk-based approach for Very Large Online Platforms (VLOPs), requiring them to assess and mitigate systemic risks. This includes obligations to assess risks related to the protection of minors, restrictions on certain advertising practices, and prohibitions of dark patterns, i.e. manipulative design choices intended to influence user behaviour.
The focus on systemic design risks is highly relevant in light of developmental psychology: given the evidence that engagement-optimised architecture and publicly quantifiable feedback can amplify adolescent developmental sensitivities, regulatory interventions should prioritise age-appropriate design standards, stricter enforcement of prohibitions on manipulative interface design, and greater transparency regarding recommender systems.
A harmonised approach at the EU level may also prevent regulatory fragmentation across Member States, an important consideration given the cross-border nature of digital services. However, legislative ambition alone is not enough. While the Commission directly supervises VLOPs under the DSA, effective enforcement still depends on the capacity to independently evaluate algorithmic systems and platform design choices, rather than relying on platforms’ own systemic risk assessments only. Ensuring that supervisory authorities have the technical resources to do so is therefore as important as the legislative framework itself.
Against this background, several researchers have begun to formulate more concrete proposals on how the DSA’s general risk-based framework could be operationalised. In this context, the Leopoldina Foundation proposed four measures at the EU level to strengthen online protection for minors.
- First, under the eIDAS 2.0 Regulation, Member States are required to provide a so-called European Digital Identity Wallet (EUDI-Wallet) by the end of 2026. This might have promising implications for age verification on social media platforms. Currently, age verification is often left to users themselves and is therefore easy to circumvent. The EUDI-Wallet could enable more reliable age verification while still complying with the strict data protection standards applicable to minors. Reliable age verification is the prerequisite for all subsequent protective measures, from phased design features calibrated to adolescent developmental stages, to restrictions on engagement-optimising mechanisms such as infinite scroll or public popularity metrics. At the same time, such age verification mechanisms may entail trade-offs for users’ own autonomy and privacy, which themselves are developmentally important aspects for adolescents. The protective gains need to be weighed against such potential digital rights costs for users of all ages.
- Second, the Foundation recommends setting the minimum age for social media use at 13 years. Given the evidence base discussed earlier, 13 as the minimum age appears arbitrary. While the evidence points to the 10-15 age range as particularly vulnerable, it does not identify a single scientifically optimal cut-off. The exact minimum age, therefore, is a question of political judgment. Yet, the more consequential question, from a psychological perspective, is how the platform experience is designed for those who do have access.
- Third, the Foundation suggests that parents require better tools to monitor and guide adolescents’ social media use. For an effective mitigation of problematic usage, there should be a simpler, cross-platform system in place that allows parents to oversee aspects such as screen time. However, from a developmental psychological perspective, adolescence is a period characterised by identity exploration and growing independence. It is therefore fundamental for these tools to ensure minors’ privacy and autonomy sufficiently, as overly controlling monitoring could work against the developmental independence that adolescence requires.
- Fourth, in line with the DSA, certain minimum standards for adolescent accounts should be defined. These could include prohibiting personalised advertising, ensuring age-appropriate algorithmic design, restricting content that may harm mental health, limiting push notifications, disabling autoplay and infinite scrolling features, and potentially allowing interactions only with approved contacts. Of the four measures, this is the most psychologically impactful, as it directly connects the platforms’ design choices with adolescents’ developmental sensitivities. Disabling infinite scroll targets the lack of stopping cues that challenge still-maturing self-regulation; limiting push notifications reduces the intermittent reinforcement patterns that heighten compulsive checking; and restricting public popularity metrics addresses the quantified social comparison that amplifies peer approval concerns during adolescence. The DSA actually already sets the legislative stage for such measures; the challenge lies in defining enforceable standards and ensuring consistent implementation across Member States.
Educational Interventions
At the educational level, research emphasises that interventions should promote digital literacy, including safe and responsible social media use, alongside socio-emotional skills such as self-regulation, emotional regulation, self-esteem, and social skills. It also emphasises the importance of mental health literacy to help young people recognise early signs of psychological distress, reduce stigma, and encourage help-seeking behaviour. Education should also include understanding the economic models of social media platforms, as well as developing media literacy skills such as identifying misinformation and verifying sources. Importantly, these competencies should be fostered not only in students but also in teachers and staff, enabling them to detect and prevent problematic social media use and to promote alternative activities such as physical exercise, arts, and music.
Translating these insights into regulation involves both the Member State and EU level, with Member States holding the primary competence in educational matters. At the national level, one concrete measure would be to restrict smartphone use in schools – as the Leopoldina report recommends for Germany, up to the equivalent of 10th grade – thereby reducing screen time during developmentally sensitive periods. At the EU level, while education remains a Member State competence, the Union has a variety of soft governance mechanisms to set an educational agenda. The European Pillar of Social Rights, particularly its first principle on education, training, and lifelong learning, provides a framework within which digital, mental health, and media literacy could be promoted. Progress can be made visible through the Social Scoreboard and the European Semester, which already track indicators such as digital skills among young people. Funding programmes like Erasmus+ and Horizon Europe can steer investments towards educational interventions related to digital, mental health, and media literacy. The Digital Education Action Plan likely provides the most direct entry point. Its second strategic priority focuses specifically on enhancing digital skills and competences and already includes actions to develop common guidelines for educators to enhance digital literacy. The framework could be extended to encompass the psychological dimensions of social media use discussed in this analysis.
Conclusion
The current debate on regulating social media for minors is often framed in binary terms: restriction versus access, protection versus autonomy. Yet, our psychological realities are more complex. To break these simplistic binaries, policymakers need to acknowledge that adolescence is not a period of dysfunction but of heightened sensitivity to social evaluation, reward, and identity formation. Social media does not create these dynamics, but defines the external environment in which they are psychologically experienced, potentially amplifying them.
A policy response grounded in psychology, therefore, cannot rely on such binaries, such as prohibition or full access. Effective regulation thus requires both structural intervention in platform design, calibrated to developmental stages, as well as educational measures that build the psychological resilience adolescents need to navigate digital environments. Both require backing by adequate enforcement and implementation capacity to have a meaningful impact.