We have been convinced for years, and with good reason, that we rely too heavily on a handful of American platforms. That the European public conversation occurs on infrastructure designed in California, governed by its laws, and fine-tuned by algorithms that reward indignation, represents a genuine problem of autonomy and democracy. Hence the increasingly robust impulse to construct European alternatives: proprietary networks, private servers, and independent rules. The desire is legitimate, although it warrants a closer examination of what is acquired when sovereignty is purchased, for changing one's skin is not synonymous with healing what lies beneath.
A social network does not become democratic by relocating from California to Frankfurt. It may be hosted in Europe, comply with European regulations, and boast verified users, while simultaneously maintaining opaque moderation, political bias, and a hostile environment for certain voices. The location of the servers dictates the governing law and the storage of data. It conveys nothing regarding whether that platform protects pluralism, moderates with transparency, or treats all published content with equality. These are two distinct questions, and we have long conflated them: we believe that by resolving the former we resolve the latter, which is not the case.
The European Union has placed technological sovereignty at the centre of its agenda, and the reasons are sound. It relies on non-European providers for more than eighty per cent of its key digital infrastructure, and this dependency constitutes a strategic risk that must be urgently reduced. There is no objection to this. The problem arises when the term “sovereignty” begins to function as if it were, in itself, a democratic guarantee. That infrastructure is European is a legal and geopolitical fact. It is not a certificate of pluralism. Yet, it is being utilised as if it were.
This misconception has an occupant. The European far right has learned to speak the language of autonomy, security, verification, and data with fluency—a language that sounds technical and civilised, thereby lowering defences. Its most effective form, and the one that causes me the greatest concern, is the seemingly neutral visualisation. A type of content circulates that presents migration as a map of dots: each dot representing one hundred people; the entire map, a stain that expands until it saturates the continent. It possesses the aesthetic of a control panel and the coldness of a statistic, and it is precisely for this reason that it functions: it converts an ideological position into an object that appears to be a fact.
The origin of these graphics warrants attention. One of the most widely circulated originates from an organisation defined by “remigration”, a term that does not merely name another migration policy. It names the proposal to reverse the presence of specific groups based on their ethnic or cultural origin, supported by the “great replacement” conspiracy theory: the notion that European populations are being deliberately replaced by complicit elites. It is important to distinguish this with precision, for confusion is the gateway through which it enters. A restrictive migration policy is legitimate in a democracy: a State may regulate borders, establish requirements, and enforce deportations with due process. “Remigration” is something else entirely—a form of demographic engineering based on the concept of ethnocultural homogeneity. Placing them in the same category, and treating the latter as if it were a rigorous variant of the former, is precisely the error that their strategy requires us to commit.
The map performs its function by omitting what is significant. A graphic that sums “accumulated persons” produces a sensation of totalising invasion, yet it erases everything that renders migration a human phenomenon: it fails to distinguish between border crossing detections and individuals—the European border agency itself warns that its figures count crossings, not individuals, and that one person may cross multiple times—nor does it distinguish between routes, asylum applications, recognised protection, or the causes that compel someone to depart from their home. It does not inform: it manufactures fear under the guise of data. And that appearance of objectivity is precisely what renders it dangerous, as it deactivates the suspicion that the same message would arouse if delivered at full volume.
None of this is new in its essence, and that is the point I wish to establish. American platforms were already accused of the inverse mechanism. Human Rights Watch documented in 2023 what it described as systemic censorship of Palestinian content on Instagram and Facebook: posts removed, accounts suspended, reach reduced. Meta qualified this and attributed it to errors and the application of its standards. It is not necessary to resolve that dispute to see the pattern: an infrastructure decides, with criteria it does not disclose, which voices are amplified and which are buried. If a European platform reproduces that same power without further oversight, we will not have changed the problem. We will have changed the address of the problem.
And here is what exacerbates the matter on new platforms. The European Digital Services Act requires transparency in moderation and mechanisms for appeals, and it provides a sound foundation. However, its most stringent obligations fall upon very large platforms, those with tens of millions of users. An emerging network may fall below that threshold and moderate with very little public scrutiny. In that twilight, moderation becomes an invisible power: it does not merely remove a post; it teaches the rest what is best left unsaid. It produces silence before it has to produce censorship. And silence leaves no trace, so it cannot even be challenged.
All of this touches culture directly, and it touches me as a part of it. Europe funds art that critically examines technology: there are public programmes and networks that support artists working on artificial intelligence, surveillance, and technical systems. It would be a bitter contradiction if, at the same time, spaces were built where that very critique, when it becomes uncomfortable, is left at the mercy of moderation without safeguards. Art works with the same elements that these platforms administer: image, memory, public emotion, and the legitimacy of what deserves to be seen. Whoever controls what appears and what is buried also controls which works and which discourses exist in public. This is not a technical question about servers. It is a question about who decides what is visible.
I am not against European technological sovereignty; I believe it is essential. I am against it being sold as a moral certificate. 'European' has become a label that promises civility, legality, and respect, and which therefore invites trust without scrutiny. But a platform does not deserve trust because of its registered office, but because of what it does: whether it publishes the criteria by which it moderates, whether it allows a decision to be appealed, whether it submits to audits, and whether it protects vulnerable voices instead of sacrificing them when they cause annoyance. Without that, the European seal is merely a seal, and behind it there may be exactly the same thing we were fleeing from.
Changing platforms is easy. Changing the regime that decides what is seen and what is silenced is the difficult part, and it is the only thing that truly matters. If the technological independence we pursue does not bring transparency, the right to reply, the protection of uncomfortable voices, and an unequivocal rejection of dehumanisation disguised as data, we will have solved nothing. We will have changed the skin of the problem, and the problem, underneath, will remain intact.
On open conversation
This text is not against any particular platform, but rather about an underlying risk: that European technological sovereignty, while necessary, might be mistaken for a democratic guarantee that it is not. Changing providers or jurisdictions does not, in itself, eradicate opaque moderation, bias, or dehumanisation disguised as data. I have endeavoured to use concepts with precision, because confusing legitimate criticism of migration with replacement ideology is the gateway through which the worst enters. If anyone wishes to intervene from the perspective of digital law, European technological policy, platform moderation, or networked cultural practice, this notebook remains open.
Sources
European Commission. Technological Sovereignty Agenda (2026); dependency exceeding 80% on non-European providers in key digital infrastructures.
Regulation (EU) 2022/2065 (Digital Services Act): transparency in moderation and complaint mechanisms; reinforced obligations for platforms with more than 45 million users in the EU.
Human Rights Watch. Meta’s Broken Promises: Systemic Censorship of Palestine Content on Instagram and Facebook (20 December 2023), and Meta's response.
Britannica, 'Replacement theory' (conspiracy theory of the 'great replacement'); Global Project Against Hate and Extremism, 'What Is “Remigration”?'.
Frontex, Migratory Map (methodological note: the data reflect detections of crossings, not unique individuals).
Creative Europe and European Media Art Platform (EMAP): European cultural infrastructures supporting critical creation regarding technology.