Every back-to-school season, the same media ritual returns: a report on teenagers glued to their phones, an expert recommending to 'limit screen time,' a parent confessing, a little ashamed, to having given in and offered a smartphone to their eleven-year-old 'because all their friends had one.' And every year, the debate exhausts itself in the same deadlock: blaming families for a problem that was designed—in the most literal sense—by engineers paid to maximize time spent on an application, regardless of what that time does to the person using it.

I want to state it from the first line, because it is the point that is systematically avoided: the problem is not our children's addiction; it is the architecture deliberately built to produce it. We don't blame a casino player for 'struggling to stop' when the slot machine has been calibrated by probabilists. Why should we continue to blame a fourteen-year-old for not resisting an infinite feed designed by the world's best attention engineers?

An Architecture, Not a Vice

The mechanism is well-documented and not mysterious. Infinite scrolling, intermittent notifications calibrated on the variable reinforcement model—the same psychological principle that makes gambling addictive—recommendation algorithms that learn in a few dozen interactions exactly what content keeps an adolescent brain glued to the screen. These are not unfortunate side effects of a neutral product. These are design objectives, measured, optimized, continuously tested by teams whose sole success indicator is time spent and daily return rate.

And this system does not target an adult capable of mobilizing mature executive control. It targets a developing brain, whose prefrontal cortex—the one that allows self-regulation and evaluation of long-term consequences—will not finish maturing until around twenty-five years of age. We have built the most sophisticated attention-grabbing machine in human history, and we have primarily pointed it at the brains least equipped to resist it.

What the Figures Cautiously Say

We must proceed rigorously here, without succumbing to moral panic. Psychologist Jonathan Haidt, in *The Anxious Generation*, argues for a mental health crisis directly attributable to smartphones. Other researchers, including Candice Odgers, dispute the robustness of this causality and point out that large population studies show only real but modest correlations.

This controversy must not serve as an alibi. In France, the parliamentary inquiry commission on the psychological effects of TikTok found that the annual prevalence of depression among adolescents rose from 2% in 2014 to 9% in 2021—a progression that coincides with the widespread intensive use of social networks. And internal Meta documents, made public in 2021, showed that the company knew that some of its products worsened body image for some of its adolescent female users, without this significantly changing its design choices.

What is striking in France is the early age of exposure. According to an Arcom study, one fifth of eleven-year-olds report having consulted social networks before their eleventh birthday, compared to only 4% of 17-year-olds at the same age—a sign that the entry age is decreasing year after year. The 2025 youth and digital barometer is even more staggering: over 90% of 13-17 year olds report regularly using social networks.

A notable fact, which should reassure us as much as it alarms us: 83% of young users surveyed by Arcom are aware of the risks they face, and 53% themselves demand more support and regulation. These are not unconscious children who need to be lectured. These are lucid adolescents, facing an imbalanced power dynamic that they want adults to correct.

Individual Responsibility as a Smoke Screen

What deeply irritates me is the consistency with which the industry has managed to shift the burden of responsibility onto families. “Activate parental controls.” “Limit screen time.” These recommendations are not wrong, but they are based on a misleading premise: that the problem is a lack of individual discipline in the face of a neutral product, rather than a design flaw in a product deliberately conceived to overcome that discipline.

This is exactly the rhetoric deployed for decades by the tobacco industry—'smoking is an individual choice'—while minimizing internal evidence on nicotine's addictive potential. We know how that story ended: with binding regulation, advertising bans, mandatory warnings. Not with an appeal to smokers' individual responsibility.

The Tentative Beginnings of Regulation

It would be unfair to pretend that nothing is moving. France this year adopted the ban on social networks for under-15s without parental consent, with platforms required to implement age verification mechanisms. The CNIL advocates a model based on trusted third parties and double anonymity: the platform receives age confirmation without ever accessing the user's full identity. These are real advances, even if the Conseil d'État has warned of the risk of friction between a national law and the European Digital Services Act framework.

But they remain patches on a system whose architecture has not changed. Age verification does not alter the functioning of infinite scrolling once the user is authenticated. Limiting nocturnal notifications does not address the recommendation logic which, during the day, continues to optimize for maximum engagement. As long as the economic model remains indexed to captured attention time, the corrective measures will remain cosmetic.

What Families Can Really Do, Without Lying to Themselves

I do not want to consign parents to total powerlessness. There are real levers: delaying the age of the first smartphone with unlimited access, prioritizing phones with deliberately limited features for younger children, establishing screen-free zones and times—the bedroom at night, the dinner table, the school commute—and above all, maintaining an open and non-judgmental dialogue about what the child sees and experiences online, rather than punitive surveillance that drives problematic usage underground.

What works best is neither total prohibition nor laissez-faire, but a form of active accompaniment: present, curious parents who are informed about the platforms their children use, capable of talking about them without immediate judgment. This is more time-consuming than a dry prohibition posted on a fridge. It is also what comes closest to an effective response.

What Must Change, And What Does Not Depend on Families

No parental strategy, however disciplined, can sustainably compensate for an architecture designed by thousands of engineers to precisely overcome this type of discipline. It is an absurd asymmetry of means to ask a single parent to counterbalance research budgets entirely dedicated to optimizing engagement.

The real shift will come the day platforms are required, by law and not merely by reputational pressure, to account for the measurable effects of their design choices on minors' mental health—just as the pharmaceutical industry must prove the safety of its products before they are marketed, not after. We would never ask a parent to single-handedly compensate for the toxicity of a poorly designed toy. Yet we continue to ask the impossible of them in the face of infinitely more pervasive and infinitely less regulated products.

Editorial Opinion

The distinction made by this opinion piece—objective isolation versus felt loneliness—is what is most often missing from public policies, which are still largely calibrated on counting contacts rather than their quality. The OrChair editorial team highlights the most counter-intuitive insight of the text: friction is not a user experience flaw; it is a condition for connection. A decade of product design has been dedicated to eliminating it; the next decade may need to be dedicated to rehabilitating it.

*Opinion. The views expressed are solely those of the author.*