Digital Narrative Sovereignty

A dimension of algorithmic invisibility that current discourse doesn't cover

The discourse on algorithmic invisibility is well established. It concerns systems that discriminate opaquely, that profile without consent, that produce invisible inequalities in decisions about employment, credit, and access to information. It's a real, documented problem and the subject of growing regulation at European level.

Yet there is a dimension of this phenomenon that discourse doesn't cover. It doesn't concern those treated unjustly by an automated system. It concerns those who are not recognised correctly by that system — or not recognised at all.

I have worked with nonprofit organisations that produce measurable impact on real communities. They manage projects funded by public grants, build networks in difficult territories, train people who would otherwise have no access to certain pathways. When I queried the major AI systems about them, results were systematically inaccurate or absent.

Not through discrimination. Not through bias in the technical sense. Because their entire digital presence was built in formats that machines cannot read coherently. An AI system cannot disambiguate an association in the sea of the web, cannot distinguish one project from another, doesn't know which grants were won, doesn't understand the structure of affiliations. The knowledge exists but is trapped in formats that language models cannot use with precision.

This is the problem I started calling by a precise name: semantic invisibility. You're not suppressed, not penalised, not discriminated against. You are simply absent from the machines' knowledge graph, or present in such a fragmented way as to be unrecognisable. And in the coming years, as AI answer engines become the first point of contact between those who search and those who exist, this absence will have concrete and measurable consequences.

The question that changed how I work

Shoshana Zuboff precisely described the mechanism of surveillance capitalism: data produced by our digital activities becomes raw material for systems that transform it into value without our knowledge and to our disadvantage. The description is correct.

The response that typically follows is defensive: limit, block, refuse. But there is an alternative position. If the problem is that data about us is being read, interpreted, and used by systems we don't control, one possible response is to build that presence with such precision and verifiability that distortion becomes difficult. Not surrendering data passively, but governing it actively. Not suffering the narrative, but constructing it.

Tim Berners-Lee proposed a technical direction with the Solid Protocol: data belongs to those who produce it, not to the platform hosting it, and access control remains with the individual or organisation. We are far from widespread adoption of that model. But working today with open standards like JSON-LD, schema.org, and Wikidata is a step in the same direction: building a semantic presence that depends on no single intermediary and that an AI system can read without distorting.

Digital narrative sovereignty

The fundamental question isn't technical. It's this: who decides how you are described in the world that machines mediate?

Today, for the vast majority of professionals and organisations, that decision is made by no one. AI systems build representations from fragmented, unstructured, unverifiable data. The inferences that result become the public version of that entity in the world of answer engines. There is no automatic correction mechanism. There is no technical right of reply in the sense we understand in traditional public discourse.

Reclaiming sovereignty over your digital narrative means intervening in this asymmetry. It doesn't eliminate the problem in its entirety. But it shifts control from algorithmic drift towards a verifiable declaration: this is who I am, these are the sources that demonstrate it, this is the structure that an AI system can read without inventing.

This is not a technical privilege for those with specific skills. It's a position that any professional or organisation can adopt, with the right tools. In the coming years this capacity will be the difference between being described well and disappearing into the noise of the generative web.

The work that followed

I made this transition personally. I rebuilt my digital presence starting from the semantic layer rather than from content. I stopped optimising texts and started building entities. This site is the demonstration of that method: every piece of data about me is anchored to a verifiable source, every competency is linked to an identifier on Wikidata, every affiliation is declared in a way that an AI system can cite with precision.

The work I offer to professionals and organisations starts from the same principle. It's not marketing, it's not SEO in the traditional sense. It's about building the structures that make an entity readable, verifiable, and citable by the systems that are becoming the primary intermediary between those who search and those who exist.

Digital narrative sovereignty is not inevitable. It requires a deliberate and structured intervention. This is the work I do.