Introduction
Today's tech giants are no longer mere market players. They have become hybrid infrastructure operators who shape the way we communicate and consume information.
This article analyzes the transition from digital monopolies to the status of quasi-institutions. You will learn how European law attempts to curb the power of Big Tech through the concept of digital constitutionalism.
The text explains why traditional regulations are failing and how instruments such as the DMA, DSA, and AI Act are transforming technology oversight into a process of systemic risk management.
Big Tech as a Hybrid Infrastructure Operator
Large technology companies are more than just monopolists. While a traditional monopoly focuses on pricing, Big Tech controls the very structure of the market and the terms of participation within it.
They act as gatekeepers, serving simultaneously as infrastructure owners, dispute adjudicators, and designers of recommendation systems. They create ecosystems in which users are partially locked in by high switching costs.
An app store is a prime example: it does not merely sell a product, but sets the technical rules for the visibility of other companies. In this way, control over network nodes grants them power far exceeding simple market share.
The Gatekeeper as Quasi-Legislator and Choice Architect
Big Tech exercises power through private ordering, creating private rules that supersede state law. According to the principle of code is law, computer code regulates behavior more effectively than legislation.
A key tool in this process is choice architecture. Through interface manipulation and nudging mechanisms, platforms shape user decisions—not by stripping them of formal autonomy, but by steering the probability of their choices.
The European Digital Markets Act (DMA) attempts to limit this. It does not merely question pricing, but asks whether a company stands at the gateway to the market and whether it leverages that position unfairly against competitors.
Big Tech as a Quasi-Institution Requiring Procedural Legitimacy
Platforms perform state-like functions: they establish norms, impose sanctions, and conduct appeal procedures. They are becoming quasi-institutions whose decisions can destroy a business or a creator's visibility.
The user's sense of helplessness stems from the diffusion of responsibility. Decisions are generated by a chain of algorithms and moderators, creating a Kafkaesque effect similar to The Castle, where there is no single point of accountability.
To remedy this, procedural legitimacy is essential. Users must have the right to a clear explanation of decisions and a genuine path for appeal, shifting oversight from individual pieces of content to risk architecture.
Summary
The struggle for digital sovereignty requires a transition toward polycentric oversight. This implies a system in which regulators, researchers, and users collectively monitor infrastructure to avoid both private and public authoritarianism.
The true challenge is not merely the regulation of AI or platforms, but the creation of mechanisms that allow us to understand and challenge the rules influencing our choices.
Freedom in the age of algorithms is not the absence of choice architecture, but the actual ability to consciously control it.
Frequently Asked Questions
Why are big tech companies more than just ordinary market monopolists?
Big tech companies are not merely producers competing for customers; they are becoming infrastructure creators, market operators, owners of participation rules, and dispute arbitrators. They serve as gatekeepers to users and designers of systems that co-create the market itself and the technical conditions for the presence of other entities within it.
How does Big Tech exercise power over the market and users beyond the traditional understanding of a monopoly?
Big Tech exercises power as "gatekeepers" who control other entities' access to the market and users, simultaneously acting as both a participant and a regulator of the environment. This power manifests through setting access conditions, designing choice architecture in interfaces, and creating private rules (code as law) that can technically restrict user behavior.
In what way can a private technological platform perform functions similar to state institutions, and where does the source of the user's sense of helplessness lie?
A private platform performs quasi-institutional functions by establishing rules, allocating attention resources, imposing sanctions, and conducting appeal procedures. The user's sense of helplessness stems from the diffusion of responsibility within a complex algorithmic and bureaucratic system, which makes decisions appear impersonal.
How does European law (DSA) change the approach to the responsibility of large technological platforms for the operation of their algorithms?
The DSA changes the approach to responsibility by moving from liability for specific acts to partial responsibility for risk architecture. Platforms are no longer treated as neutral containers for content; instead, they must assess and mitigate systemic risks generated by their algorithms, interfaces, and business models.
What does the actual power of large technological platforms consist of, and how does law (e.g., DMA) attempt to limit it?
The actual power of technological platforms stems from their position as essential infrastructural nodes, allowing them to control information flows and set visibility conditions for the decisions of other network participants. Law, including through the DMA, limits this power by prohibiting gatekeepers from unfairly exploiting their control over infrastructure and from giving preferential treatment to their own services at the expense of dependent entities.
How do European regulations such as the DSA counteract Big Tech's total control over behavioral data?
The DSA counteracts Big Tech's control by introducing a model of regulated polycentrism, which imposes obligations on platforms regarding transparency, auditing, and risk management. A key tool is ensuring qualified researchers have access to data, enabling independent measurement of system effects and the democratization of observability regarding their impact on society.
How does the AI Act regulate artificial intelligence, and what risks are associated with the process of implementing these regulations?
The AI Act regulates artificial intelligence based on risk categories and application functions, introducing transparency requirements and oversight for general-purpose AI models. A risk associated with this process is the so-called 'regulatory moat,' a situation where high compliance costs become a barrier to entry for startups and reinforce the dominance of the largest players.
How is the European Union attempting to regulate Big Tech, and why is traditional law insufficient in this case?
The EU regulates Big Tech through legal acts such as the DMA (limiting gatekeeper practices), the DSA (regulating platform processes and risks), and the AI Act (liability for AI systems). Traditional law is insufficient because it assumes the stability of the subject being regulated, whereas technology platforms can change their code daily.
How can fairness and control be ensured in AI systems that make billions of decisions too quickly for traditional legal procedures?
It is necessary to move from controlling every single decision to a 'due process by system design' model, combining law with safety engineering. Fairness is ensured through the design of audits, error testing, transparency of the governance model, and effective appeal procedures in high-stakes cases.
Why is traditional law insufficient for controlling Big Tech, and what should effective AI regulation look like?
Traditional law is insufficient because platforms integrate multiple roles (e.g., acting as both a bank and a medium), and technology evolves faster than legal procedures, creating so-called epistemic latency. Effective AI regulation requires technologically neutral categories and the transformation of the regulator into a 'learning institution' that utilizes tools such as regulatory sandboxes and algorithmic audits.
How can the power of Big Tech and AI be effectively limited to avoid replacing private authoritarianism with state authoritarianism?
Effectively limiting the power of Big Tech and AI requires implementing a polycentric system of control, where regulators, courts, independent researchers, and users can mutually challenge platform decisions. It is essential to build public human capital and technical infrastructure, as well as ensure data interoperability, which increases users' freedom of exit. To avoid state authoritarianism, digital constitutionalism must simultaneously limit both private and public power.
