From Voluntary Servitude to Reconquest
Towards a doctrine of public digital power
We are in 2026, and the awakening is painful. For more than a decade, reports have been piling up, parliamentary investigative committees have been following one another, and findings of terrifying lucidity have been drawn up. Yet, in ministries as in the management committees of our industrial flagships, the same torpor persists. The recent hearing of Henri Verdier before the national representation should have had the effect of an electric shock. It painted the clinical panorama of a nation and a continent undergoing accelerated technological vassalization.
As an entrepreneur in the sector and a consultant familiar with the arcane workings of public affairs, I can only rage at this suicidal wait-and-see attitude. The digital illiteracy of our decision-makers is no longer a simple technical gap; it has become a major political fault, a direct renunciation of our sovereignty and, ultimately, of our democratic model.
The diagnosis is before our eyes, implacable. It is high time to abandon semantic incantations to confront the brutality of real power relations.
The Invisibility of infrastructure and the privatization of the World
The French public debate, when it deigns to take an interest in digital technology, focuses with an almost neurotic obsession on the purchase of software licenses or cloud hosting. This is looking at the world through a keyhole. Our dependence is of an infinitely deeper, more systemic nature. It strikes the hardware that we no longer know how to produce. It strikes our cybernetic capabilities, where, despite the excellence of our security agencies, we struggle to anticipate the offensive capabilities of foreign states.
Even more serious, this dependence strikes critical physical and logical infrastructures. How can we tolerate the idea that the vast majority of new global submarine cables are now laid, financed and controlled by a handful of hegemonic American technological players? We are witnessing, in deafening silence, the privatization of the physical infrastructure of the global network.
Similarly, entire sections of our economy rely on fundamental data that seem free and acquired to us, such as geodesy, spatial positioning or cartography. These de facto commons often rely on extra-European infrastructures, sometimes under-funded by their own creator states, and which could collapse or be brutally monetized overnight.
The major information access services, social networks, search engines, these famous "gatekeepers", today exercise a domination that extends far beyond the technological sphere. They impose technological standards and, inevitably, ideological standards. Europe's inability to invent its own path in the field of artificial intelligence, refusing to exit the paradigm of the race to gigantism of large language models propelled by titanic data centers, is glaring proof. We abdicate our vision of the world to embrace that of Silicon Valley, forgetting that other paths, championed by European researchers or nations like India, advocate for distributed, frugal and open models.
The digital proletarianization of French Industry
It is imperative to understand the nature of value capture at work. The drama of delivery platform drivers, whose lives have been transformed into an algorithmic hell dictated by incessant data flows, is only the vanguard of a macro-economic phenomenon. These workers, isolated from any human hierarchy, see their remuneration and working conditions reevaluated in real time by an artificial intelligence optimizing the profit of the central platform.
If we are not careful, our large companies and our administrations are about to suffer the same fate. This is the threat of the "uber-deliveroo-ization" of the French economy. From the moment digital technology, artificial intelligence and extra-European critical infrastructures interpose themselves at the heart of our value chains, they acquire the power to permanently redirect margins to their sole profit.
Our industrial flagships, whose real added value will increasingly reside in data and algorithms, risk being reduced to the rank of simple subcontractors for infrastructure monopolies. Changing the nationality of a technological tyrant does not solve the problem; trading a foreign monopoly for its European equivalent will improve neither our democratic health nor the freedom of our markets.
Extraterritoriality and Geopolitical Vulnerability European angelism must cease in the face of the legal and political arsenal of our allies and rivals. Our technological dependence makes us structurally vulnerable to espionage, institutionalized by foreign legal frameworks that authorize access to data under their jurisdiction, regardless of where it is physically stored.
Even more violently, this dependence exposes us to unsustainable unilateral retaliation measures. The case of this French judge of the International Criminal Court remains the terrifying symbol of our impotence. For having displeased the American president in the exercise of his functions, this man was struck by an executive decree initially designed to fight international terrorism. Overnight, cut off from payment infrastructures and American digital services, he lost access to transport, entertainment services, and even everyday French services, so intertwined are foreign payment APIs in our ecosystem. An arbitrary, unilateral decision, bypassing any judicial procedure, was enough to socially and economically extinguish a European citizen. What do we do about this? Europe nevertheless has blocking mechanisms that are never activated, preferring to let its companies sink into fearful over-compliance with foreign injunctions.
Absolute independence does not exist, China and the United States regularly prove this to us in their trade wars over rare earths or semiconductors. The real question is that of the balance of power. Yet today, we are not in a position to speak as equals. We are treated as a marginal adjustment variable.
The cultural incompetence of the purchasing State
The heart of the problem, the tumor that eats away at our capacity for resilience, is above all cultural. Observing the public administration and a large part of the private sector, it is clear that digital technology is still perceived there as an ungrateful support function. The information systems management is relegated to the same rank as real estate management or collective catering, when it should be the eminently strategic command post, sitting on the executive committee or directly influencing ministerial decisions. We have transformed our administrations into huge purchasing centers, populated by civil servants who no longer know how to design, who no longer know how to code, who no longer know how to test. When a State no longer knows how to do, it no longer knows how to buy. It finds itself at the mercy of service providers who sell it specifications, absurd specifications, support for supplier selection, and ultimately monolithic projects spread over five to ten years. These projects, conceived as bureaucratic cathedrals for military pay or human resources management, systematically collapse under the weight of their own complexity priced at millions of euros. The figures are humiliating. When the State spends four billion euros per year on IT for its ministries – half of which evaporates in the purchase of North American proprietary licenses – a single large French bank spends twice as much. This chronic underinvestment is the symptom of a State that trembles before the digital revolution, which delays the moment of action for fear of the inevitable failure induced by its own sclerotic processes. It is appalling to hear IT directors from both the public and private sectors justify maintaining obsolete proprietary solutions, costing millions and concentrating the majority of security flaws, out of simple pusillanimity in the face of free and open alternatives. Courage has deserted the ranks of those who should protect our decision-making architecture.
The platform State model and the wisdom of digital public infrastructures
Yet other paths exist, marked out by spectacular successes that we refuse to industrialize. The notion of Platform State, theorized for more than a decade, demonstrates that another approach is possible. Instead of building gigantic closed systems, the State must design bricks, APIs (application programming interfaces), fluid and secure modules allowing ecosystems to innovate on a common base. India's example is in this regard a lesson in humility and strategy. Faced with the threat of their market being captured by monopolies like Uber or Google, Indian architects did not seek to create a competing public company, nor to lock themselves into purely punitive regulation. They built Digital Public Infrastructures (DPI). By imposing open standards for identity, payment and geolocation, they guaranteed that the logical infrastructure remains under public control, while letting the market proliferate on top. The result? An illiterate rickshaw driver can operate via a local application developed for a few thousand euros, freeing himself from the 40% tax levied by intermediary platforms. The Indian State intervenes not to prevent innovation, but to prevent the takeover of a two-sided market by an unfair intermediary. This is a doctrine of absolute finesse: leave innovation to the free market, but fiercely protect the logical rails of the economy. These infrastructures cost ridiculously little, because they involve architecture design and standards, not titanic data centers. In France, when public engineering has been granted trust, it has been able to deliver gems like FranceConnect or the State's secure messaging (Tchap), developed on open source bases for a cost per public agent that is derisory compared to predatory market offers. These victories of the agile method, internalized development and free software demonstrate that fatalism is not appropriate. French technical excellence exists, we just need to free it from the shackles of blind buyers.
The lever of commons and the illusion of the protectionist bubble
The quest for sovereignty must not sink into the fantasy of absolute autarky or blind national preference that would condemn our companies to scrape by in a non-competitive bubble, selling solutions twenty times too expensive to a captive public buyer. Reasonable autonomy passes through diversity, through the requirement of loyally competitive markets, and above all through massive investment in Digital Commons. The existence of free software, open standards, projects like OpenStreetMap or Wikipedia constitutes a formidable arsenal of sovereignty. These commons do not give anyone absolute control of the situation, but they concretely prevent a monopoly from seizing them. They are the foundation on which we must base our defensive doctrine. Yet the State continues to largely benefit from these ecosystems without funding them commensurate with their criticality. Saying "thank you for making us free" is not enough in the face of American giants who do not hesitate to massively fund these projects to better absorb them from within, as demonstrated by the gradual integration of certain free internet actors into the orbit of dominant search engines.
Europe, regulatory battleground and libertarian lobbying
Faced with this asymmetry of power, the European level remains the only relevant critical mesh. The Digital Services Act (DSA) and the Digital Markets Act (DMA) are elegant, fundamental texts that precisely address content regulation and market concentration by establishing a logic of compliance and accountability.
But we sin through guilty naivety about the brutality of the response from the other side. Europe swarms with lobbyists whose objective is not necessarily to whisper counter-truths, but to drown public power under complexification. The recent "AI Act", a bureaucratic monster of hundreds of pages whose essence no one grasps anymore, is the sad trophy of this exhaustion strategy.
Even more alarming is the ideological infiltration operated by the libertarian fringe of Silicon Valley. An assumed ideology, theorized more than twenty years ago, postulating that anti-regulation is the absolute condition for innovation and that only monopolies deserve to exist. When tutelary figures of this industry, close to circles of power in Washington, publicly threaten to campaign for NATO withdrawal if Europe dares to apply fines to their social networks, we change dimension. This is no longer about commerce, it is about pure geopolitical coercion. Are we ready, politically, to withstand this shock? Will we have the political courage to enforce our laws at the price of potential transatlantic reprisals?
It would be false to believe that European or French public decisions are vulgarly corrupted. The evil is more insidious. It resides in these "evening visitors", these representatives of large consulting firms and hegemonic actors, who permanently whisper simplistic thinking frameworks in the ears of our leaders. They instill the idea that entrusting our sovereign data to foreign black boxes, under the fallacious pretext of immediate efficiency, is the only reasonable path.
The scandal of public imprudence
The consequences of these whispers are tragic. The massive data leaks that regularly strike the public sector are only the reflection of a surface of risk exposure built on unheard-of conceptual weakness. How to justify hosting the national health data platform with an extra-European actor via a bypass of classic tenders? How to accept, without trembling, that our intelligence agencies make themselves absolutely dependent on foreign technology companies whose economic model consists precisely in making themselves indispensable, then dictating their pricing grids and their operating conditions, under the implicit threat of unplugging the State's capacity to exercise its sovereign power?
The illusion of the "trust bubble" around servers disconnected from the network in no way erases the dispossession of know-how. When a commercial or political conflict arises, we will no longer know how to operate without their algorithms. We are outsourcing our cognitive sovereignty.
An action plan for the coming decade
It is not too late, but the time for lukewarmness is over. Europe has not made the bet on self-sufficiency, but it can still build a viable balance of power. For this, our political and economic leaders must impose a new doctrine of action, strict and implacable: Re-internalize strategic public engineering: The State must relearn how to do. We must stop funding an army of specification buyers to massively recruit developers, atypical profiles, supporters of agile methods. Public digital procurement must become "intelligent", stop seeking the illusory guarantee of monolithic decennial contracts, to advance by iteration and rapid prototyping.
Impose reversibility and independence by contract: No public contract signature with a major cloud or AI supplier should be validated without the integration of a strict reversibility clause. Let us demand the realization of blank migration tests before any deployment: if the State cannot leave the solution and recover its data in fifteen days, the contract is void. Each daily decision must be evaluated in light of a single question: "Does this option increase our degrees of freedom?" Build the Platform State and fund the Commons: Let us abandon siloed information systems. Let us build robust, secure, documented APIs. Let us stop wanting to operate final services where the ecosystem can take care of it, but let us regain absolute control of the technical regulation layer. Let us invest massively, at the European scale, in the foundations of the open internet, so as not to leave this domain of struggle at the mercy of foundations satellitized by the GAFAMs. Support European industrial artificial intelligence: Let us turn away from the deadly fascination with generative AI for the general public and the irrational energy consumption of pharaonic data centers financed by opaque capital. Europe's future lies in AI applied to industry, using non-public sectoral data to optimize our production chains, our ecological transition and our material sovereignty. Exit regulatory naivety: Let us understand that regulation is not the enemy of innovation, provided it is elegant and focused on protecting democratic models. Let us demand that Brussels maintain the line of firmness on the application of the DSA and DMA. Let us stop giving in to the blackmail of technological delay waved by those who have a precise interest in us not setting any rules.
Digital sovereignty is the contemporary name of democracy
The message is clear, and it is addressed to you, decision-makers, ministers, general directors: digital is not a cost center to outsource. It is the central nervous system of our nations. If the sovereign people lose the capacity to decide on the protection of their data, access to their information, and the rules governing their economy, there is no more democracy. The history of digital is not frozen. The giants who dominate us today may very well collapse tomorrow in the face of new technological paradigms. The darwinism of innovation is permanent. But we can only seize these opportunities for reversal if we cultivate in ourselves this will to power, this culture of concrete independence. Prepare the following battles. Do not miss the next logical architecture. Stop begging for permission to regulate from those who wish our intellectual dissolution, and start building the load-bearing walls of our own edifice. Our survival as a free continent depends on it.
