Age Verification on Social Media in the Light of Libertarianism

This piece is published in response to the following article in the Greek news outlet PRÓTO THÉMA: https://www.protothema.gr/greece/article/1893385/pos-tha-leitourgei-o-koftis-sta-social-media-gia-paidia-kato-ton-15-eton-live-i-parousiasi-ton-metron/ and to a later report by the same outlet: https://www.protothema.gr/greece/article/1893507/bloko-se-instagram-facebook-kai-tiktok-gia-paidia-kato-ton-15-eton-oi-platformes-tha-eleghoun-tin-ilikia-ton-hriston (both in Greek).

Part 1

Protecting children from the problematic use of social media is admittedly a serious matter. For precisely that reason, it is not enough for a regulation to offer an image of resolve: it must show that it reduces harm without a disproportionate cost to freedom and privacy.

The Greek bill, to be voted on Thursday 22 October and to apply from 1 January, as presented in the PRÓTO THÉMA report above, introduces an age-based exclusion from the social-media services it covers, with no exception for parental consent and no protection for accounts that already exist. The crucial question is not only which children will be excluded, but which control mechanism all other users will have to accept as well.

Precision is needed here: age verification does not necessarily amount to disclosing one’s identity. A properly designed digital credential can prove that the age threshold is met without revealing a name or a date of birth. The claim that “everyone will hand over their identity to the platforms” is therefore not substantiated. What is substantiated, however, is that age assurance potentially concerns adults too: to exclude those who are not eligible, you need an accepted basis for distinguishing those who are.

A Misesian critique begins with the distinction between intentions and results. The existence of risks does not prove that this particular ban is effective. Success is not measured in closed accounts, but in better sleep, less problematic use, and greater safety. One must also take into account children moving to less safe services, wrongful exclusions, and the cost of compliance, which may fall disproportionately on smaller providers.

There is also the risk of escalation: circumvention of the restriction may be used as a justification for stricter controls and for extending the ban. This is not an inevitable development. It is a reason to set clear limits, so that every failure of the intervention does not automatically become an argument for more intervention.

From the standpoint of Murray N. Rothbard, the matter is, in addition, a matter of coercion. The terms that a private service chooses are one thing; a state-imposed exclusion that you cannot escape by changing provider is another. The absence of an exception for parental consent means, moreover, that the state does not merely support parents: it replaces their judgment on this particular access. Children have rights of their own, but this calls for weighing their protection against their access to information, their expression, and their social participation — not for bypassing the debate.

Zero-knowledge proofs can offer real protection. On their own, however, they do not guarantee the privacy of the whole system. Who records the use of the credential? Can different accounts be linked? Who revokes access, and how are errors corrected? “No data at all” is not an accurate description when even an age indication is transmitted.

This is where the Orwellian warning is reasonable. An access-control infrastructure can accustom society to communicating under mandatory certification. If, in addition, protective language is used to brush objections aside, a legitimate aim becomes a shield against democratic scrutiny.

Priority should go to the least intrusive means: support for families, digital literacy, easy-to-use restriction tools, and action against specific harmful practices. Every mandatory mechanism requires independent audits, data minimization, a right of appeal, and genuine review.

Protecting children is a duty. It is not a blank check for controlling everyone.

In the light of libertarianism, the crucial line lies between protection against specific violations of rights and compulsion for the supposed good of the individual. Dealing with fraud, threats, and the exploitation of minors is not the same as the preventive exclusion of an entire age group, regardless of how they use the services and of parental consent. That some uses are dangerous is not enough, by itself, to legitimize banning all of them. The bill must therefore justify why targeted action against harmful conduct and support for families are not sufficient.

Libertarianism does not presuppose infallible parents or innocent platforms; but it denies that their failings automatically justify the state substituting its choices for theirs. Children are bearers of rights, not the property of either the family or the state. A policy consistent with this principle would strengthen responsible guardianship, voluntary protective solutions, and accountability for specific violations, without taking for granted a mandatory age-assurance mechanism as a precondition for everyone’s participation.

The burden of justification lies with the one who imposes the restriction, not with the one who asks to remain free.

Part 2

The crucial point in the new social-media bill is not only who will be excluded. It is how everyone else will prove that they are allowed to take part.

According to the later PRÓTO THÉMA report (see above), platforms will be required to verify age by appropriate, proportionate, and reliable methods. Gov.gr Wallet and Kids Wallet are mentioned as indicative solutions: they will transmit an “age token,” that is, an indication of whether the threshold is met, without revealing the date of birth.

This distinction matters. Certifying an age attribute is not necessarily the same as revealing one’s identity to the platform. The article does not say that everyone will upload an ID, undergo a biometric check, or that the state will know every time they log in. The report does not, however, describe the full process of issuing and using the tokens, so that we would know who processes which data at each stage.

The assurance “we do not give away any personal data at all” is therefore too absolute. The age indication is information about the user and, insofar as it is linked to an identifiable account, it may constitute personal data. The right aim is to minimize data, not to make it disappear by choice of words. And the fact that platforms already collect a great deal of information does not automatically justify a new mandatory processing.

The essential questions remain: does the issuer of the token know where it is used? Can different uses of it be linked? Are logs kept? What happens when an adult has no accepted means of verification, or is wrongly excluded? The absence of a name from the message the platform receives is an important safeguard, but it does not prove, by itself, that the whole process is untraceable.

Here the question of freedom arises. A ban for children creates a need for age assurance that may also concern adults. Access to the services covered acquires an additional, legally imposed precondition. Even if it works anonymously, it remains a restriction that needs its own justification. Protecting the privacy of a check does not answer the different question of whether the check is necessary.

From the standpoint of libertarianism, moreover, delegating enforcement to private companies does not turn a state obligation into a voluntary choice. At the same time, the threat of sanctions may create an incentive to over-comply: when age is uncertain, a platform may prefer exclusion or extra verification demands. That is why we need not only obligations to check, but also safeguards against unjustified exclusions.

The Orwellian analogy must be stated here with the same precision. A mechanism of universal surveillance has not been demonstrated. There is, however, the risk that we grow used to communicating under a mandatory eligibility check, and to infrastructures that — if their design and future decisions permit — may serve wider restrictions in the near or more distant future.

“We do not play with children’s health” is entirely respectable as a statement of purpose. It is not an answer to the objections about the means being used. Nor does a majority in an opinion poll prove necessity or proportionality.

Before implementation, we need public technical specifications, independent audits, safeguards against linking uses, accessible alternatives, and an effective right of appeal. Above all, we need proof that the benefit to children justifies the restriction.

A free society is not content with the assurance that control serves a good purpose. It demands to know why it is needed, how it is limited, and who controls the one who exercises it.

Praxeologos