Dopamine, algorithms and minors: social trials begin in the US

Teenagers seem more vulnerable because their brains are still under construction and because, at that very stage, they begin to measure their own worth by looking up to peers. Social offers a continuous stream of micro-rewards. Every swipe is a promise. Every notification a reminder. Absence, on the other hand, becomes aversive. The fear of missing out makes offline uncomfortable, almost painful. The result is a cycle of reinforcement that drives one back to the screen as a defence mechanism.

An increasing correlation between intensive use of social media and signs of psychological distress has been measured. The key word, however, remains 'correlation'. The definitive causal evidence is still subject to scientific debate. And this is where the trial gets interesting: the jury will have to decide whether the design of the platforms crossed a threshold of liability, even in the absence of granitic causal evidence. In other words, whether the architecture of the digital experience can be considered a risk factor per se.

In recent years, the debate on minors and social media has changed profoundly. First the users were to blame, then the parents. Today the focus shifts to the platforms. No longer just what is shown, but how. Not just content moderation, but experience design. It is the transition from editorial control to health-by-design.

The experience of New York State, often cited as a regulatory laboratory, fits into this space. The SAFE for Kids Act, signed into law in June 2024, directly intervenes in the design of platforms, limiting addictive algorithmic feeds and banning nightly notifications to minors without verifiable parental consent. It is a structural law, but still suspended pending implementing regulations.

The approach of S4505/A5346, signed into law in December 2025 and effective from 2026, is different. Here, protection comes through mandatory, clear, non-skippable warnings, based on scientific evidence updated by the health authority. Social is treated as a potentially harmful product for mental health. The message is direct, almost brutal. The limits, however, are obvious: the warnings inform but do not change the architecture that generates addiction, and the sanctions envisaged are contained, easily absorbed by global platforms. The cultural signal is strong. The economic one much less so.

In Europe, the ground is ready for a further step. Binding guidelines could clarify that recommender systems designed to maximise minors' online time exploit age-related vulnerabilities, easing the burden of proof for authorities and shifting it to service providers. In the audiovisual field, protection could extend from content types to modes of enjoyment: autoplay, repetition, absence of interruptions. Finally, in terms of soft law, work could be done on true health-by-design standards, also involving health authorities.

In short, the trial that opens today in Los Angeles will not decide the future of social networks. But it could change the way we look at their past. And when a court starts questioning not about content, but about attention design, the echo goes far beyond the courtroom. Even beyond the screen.