# Technofeudalism: platform ownership, economic rents and the struggle over digital sovereignty

The term technofeudalism captures a dispute over the organization of digital capitalism. Yanis Varoufakis argues that large platforms have created a new form of ownership and command through what he calls cloud capital: infrastructures that direct economic behavior while appropriating rents from businesses and unpaid activity from users. Evgeny Morozov challenges the notion that these developments amount to a new mode of production, arguing that monopoly, state protection and rent extraction have long belonged to capitalism. Cédric Durand offers a related but distinct interpretation of a movement toward rentier dependence and the appropriation of value through proprietary infrastructures. [1][2][3]

The dispute has empirical consequences. If firms operating marketplaces, application stores, advertising systems and cloud computing are mainly capitalist companies with unusually strong monopoly positions, competition law, labor rights and regulation may explain their operation. If the relationship has become qualitatively different, it may require a deeper account of markets organized within private jurisdictions that control access, rankings and contractual terms. Neither position can be verified by metaphor alone.

Historical feudal relations were linked to land, jurisdiction, coercion and obligations of status. Contemporary platforms operate through private contracts, algorithms, network effects and industrial infrastructures within sovereign legal orders. The comparison is valuable only when it identifies concrete similarities in dependence and rent, while recognizing the substantial legal, technological and historical differences. Its strongest contribution may be to direct attention to ownership and economic power that ordinary measures of software-sector innovation can conceal.

## 1. Feudalism as a historical social arrangement

European feudalism was not a uniform institution across time or territory. Landholding, military obligations, local jurisdiction, taxes and peasant labor relations differed between kingdoms and centuries. Manorial institutions and wider aristocratic orders connected access to productive land with political authority. In some settings, peasants faced restrictions on movement and inherited obligations to landlords. Control over agricultural surplus could rest on legal and coercive privileges rather than on contractual exchange among formally equal market participants.

Historians distinguish feudal tenure, seigneurial rights and agricultural labor systems because they did not always coincide. A nobleman might exercise jurisdiction without directly operating every farm; commercial markets could develop alongside coerced obligations. Simplifying the entire period to a lord charging rent obscures the distribution of actual rights and the evolution toward wage labor and commercial agriculture.

Contemporary platforms are sometimes called digital estates because they organize spaces where others do business. But vendors on a marketplace generally remain separate legal entities, consumers have formal citizenship rights and employees are not tied to hereditary land tenure. The analogy must therefore focus on specific mechanisms: dependence on an infrastructure that participants cannot easily replace, private rulemaking over access and the appropriation of recurring income from that dependence.

## 2. The persistence of capitalist property and competition

Capitalism is ordinarily described through private ownership of productive assets, wage labor, investment for profit, credit and competition among enterprises. None of these features excludes monopoly. Railroads, telecoms, electric networks and patent portfolios have historically created concentrated control within capitalist systems. States have routinely enforced property rights and sometimes granted concessions or regulatory protections that reinforce dominant firms.

Digital platforms remain financed through corporate equity, debt and retained earnings. They employ workers, buy equipment, fund research and compete for commercial customers. Firms that operate within an app store or online marketplace also typically face costs, supply labor and sell products. These observations make it difficult to establish that capitalist production relations have disappeared.

Yet the conventional description may be incomplete if it treats the platform solely as another ordinary supplier. A company controlling payment processing, advertising, distribution and search visibility can establish rules that shape the market itself. Competition among sellers then occurs within an arena governed by a third party able to change commissions or rankings. Such authority is a distinct source of bargaining power even if the broader economy remains capitalist.

## 3. Economic rent from Ricardo to contemporary industrial economics

Economic rent has several intellectual meanings. In David Ricardo's classical explanation, rent relates to differences in land productivity and location, under conditions where better land becomes scarce. Karl Marx analyzed ground rent in relation to property and capitalist production. Modern economics often defines rent as a payment above the minimum needed to keep a resource in its present use. These traditions are not identical, but each directs attention to returns that depend on constrained access to something economically valuable.

An online platform may create a genuinely efficient service and also exercise market power. Search, fraud prevention, payments and logistics involve real costs. A commission can pay for those services. A portion of the same commission may nevertheless be sustained by sellers' inability to reach buyers elsewhere. Economically separating the two elements requires evidence on pricing, incremental costs, substitutable distribution channels and contractual restrictions.

A firm can simultaneously earn ordinary profit, economic rent and returns attributable to innovation. Treating rent as the entirety of platform revenue exaggerates the critique; ignoring it entirely understates monopoly power. The crucial question is how much a platform can increase charges or change participation conditions without losing enough customers to make doing so unprofitable.

## 4. Varoufakis and the notion of cloud capital

Varoufakis argues that cloud capital is not simply software or physical servers. It includes machines, algorithms and networks trained through interaction with users, allowing owners to direct the environment in which choices are made. Platforms suggest products, order information and structure social and commercial encounters. In his interpretation, this creates a form of ownership whose power extends beyond producing goods for sale in a conventional market. [1]

The notion of cloud rent describes payments made by producers who must gain access to a digital domain controlled by a platform owner. Varoufakis calls some dependent entrepreneurs cloud vassals and describes unpaid user contributions to platform value as cloud serfdom. Those expressions are explanatory analogies, not legal classifications of users as medieval peasants. They emphasize asymmetric control and the privatization of markets.

The strong claim is that dominant intermediaries have replaced a growing share of open market exchange with privately governed digital spaces. The weakness is determining how extensive that displacement actually is. Online retail and cloud infrastructure are important, but physical manufacturing, ordinary retail, public services, banking and competitive professional markets remain substantial. Establishing a new mode of production requires more than showing that some intermediaries have become powerful.

## 5. Morozov's criticism: capitalist monopoly is not new feudalism

Evgeny Morozov criticizes technofeudal arguments for presenting familiar capitalist dynamics as evidence of a historical break. Intellectual-property restrictions, predatory business strategies, state-supported monopolies and unequal bargaining relations existed long before digital platforms. Treating them as a return to feudalism can obscure their origins in capitalist accumulation and the institutional choices that permitted concentrated ownership. [2]

His critique also carries political implications. If society has entered an entirely new economic order, measures addressing monopoly power, labor exploitation or corporate governance might appear secondary. Yet those measures could be crucial precisely because platform activity remains rooted in financial investment, employment and profit-seeking. The argument against technofeudalism is not that platforms are benign, but that their power may be better explained with established categories.

A serious response must identify what changed qualitatively. Scale alone is not enough, nor is revenue earned from intellectual property. More persuasive evidence would show that an increasing proportion of production cannot be organized without compulsory participation in privately administered digital infrastructures, and that these infrastructures enjoy authority analogous to economic jurisdictions. Such evidence remains contested and varies sharply by sector.

## 6. Cédric Durand and rentier regression

Cédric Durand connects digital concentration with broader developments in finance, intangible assets and capitalist property. His work emphasizes the expansion of returns tied to ownership over infrastructures, intellectual property and organizational dependencies. It is less dependent on one analogy between a particular app store and a medieval estate than on the possibility that economic accumulation has become increasingly centered on extracting payments from controlled access. [3]

Digital assets permit new forms of enclosure. Data collections, algorithms and customer networks can be made difficult to replicate even when the underlying information is not naturally scarce. Legal rules, exclusive contracts and technical incompatibilities help maintain control. A firm that operates a central marketplace can receive revenue from other companies' sales without directly manufacturing their products.

The empirical difficulty is disentangling innovation rents, monopoly rents and payments for real intermediary services. An integrated system may reduce transaction costs at the same time that it raises switching costs. Financialization may intensify pressure to monetize dependence, but it does not prove that production has ceased to matter. Durand's framework is therefore most informative when connected to measurable ownership structures and contractual practices.

## 7. Digital platforms as privately administered markets

Online marketplaces are not neutral locations where exchange merely happens. Operators establish eligibility criteria, payment rules, dispute procedures, search rankings, advertising access and fee schedules. Sellers can compete intensely with one another while possessing limited influence over the rules defining competition. The platform's ability to amend terms introduces a relationship of dependence that conventional models of independent traders may understate.

The same infrastructure can deliver major benefits. A common search interface reduces customers' discovery costs, payment systems limit fraud and logistical services create economies of scale. These benefits explain why sellers choose to participate even when commissions are significant. The existence of value-creating intermediation is not incompatible with abusive market power.

Assessing the balance requires examining sellers' alternatives, the portion of sales dependent on the platform and changes in fees relative to the cost of providing services. Ranking preferences for an operator's own products, restrictions on alternative prices and access to commercially sensitive data may further shift competitive conditions. The analytical issue is governance of exchange, not merely the size of a technology firm.

## 8. Platform labor and algorithmic management

App-mediated work demonstrates how digital systems can reorganize employment and bargaining power. Dispatch, performance measurement, customer ratings and payment rules can be enforced through software. Workers may enjoy some flexibility in choosing hours, but platforms can also influence availability of work, income volatility and access to appeals. The distinction between an independent contractor and an employee becomes controversial when practical direction is exercised algorithmically.

Economic risks may be shifted to workers who provide their own vehicles, equipment and insurance. Gross earnings are not equivalent to net compensation after those costs and unpaid waiting time. Ranking or suspension decisions can have major consequences for livelihood even when the platform describes the relationship as a commercial partnership.

Algorithmic control is neither inherently feudal nor merely a neutral efficiency tool. Similar forms of supervision exist in warehouses, call centres and other capitalist workplaces. Its significance lies in granular measurement, opacity and the difficulty of collective bargaining against a system whose criteria are not transparent. Labor protections should respond to observable powers and material outcomes rather than rely exclusively on historical labels.

## 9. Data, attention and uncompensated activity

Users generate information through searches, clicks, location traces, social interactions and uploaded content. Platforms can use those signals to improve products, target advertising and develop models. Varoufakis describes certain forms of user activity as unpaid labor feeding cloud capital. This highlights the dependence of digital businesses on information produced beyond conventional paid employment.

The analogy has limits. Enjoying a social service while leaving data traces is not always equivalent to performing productive labor under employment relations. Economic value can emerge jointly from infrastructure, software, user participation and advertisers. The question of remuneration depends on rights, consent, market power and the purposes for which information is processed.

A sharper analysis examines what users can understand and control. Can they refuse collection without losing essential access? Can information be ported to another provider? Do terms allow the operator to alter downstream uses? Whether one calls these practices exploitation or unequal contractual exchange, they raise questions about informational autonomy and the distribution of benefits from digital networks.

## 10. Network effects and winner-take-most dynamics

A network effect occurs when a service becomes more useful to a participant as the number or quality of other participants increases. A marketplace attracts buyers when many sellers are present, and sellers when many buyers participate. A messaging service becomes more useful when its users' contacts are accessible. These feedback mechanisms can favor large incumbents and make entry difficult.

Not every digital service has the same network effects. Users can belong to several platforms, and specialized entrants may compete on quality or price. Multi-homing can reduce dependence; exclusive contracts and incompatible data formats can intensify it. Economies of scale in computation are related but not identical to network effects, since large fixed costs can lower average costs even without direct interaction among users.

Market concentration is consequently an outcome that needs explanation, not a law of technology. Regulation can affect interoperability and contestability, while product innovation can create viable competitors. A high market share may reflect service quality, exclusionary conduct or both. Identifying the mechanism is more useful than assuming that all concentration validates technofeudalism.

## 11. Cloud computing and the cost of exit

Cloud providers supply computing, storage, databases, networking and software development services. Businesses benefit by avoiding the upfront cost of building and maintaining every system themselves. Yet long-term use can make migration difficult. Proprietary interfaces, data-transfer fees, contract discounts and specialized workforce skills affect the cost of changing suppliers.

The United Kingdom's Competition and Markets Authority investigated cloud services and published its final decision in July 2025. Academic work has separately examined how data egress charges may affect dynamic competition in cloud computing. Such material provides concrete grounds for studying dependence instead of assuming every customer relationship is inherently coercive. [4][5]

A customer can formally retain the legal right to leave while facing a costly technical transition. This is a genuine asymmetry, though not a revival of serfdom. The remedy might involve interoperability standards, clearer contract terms, pricing scrutiny and architectural choices such as portable workloads. Whether regulation lowers overall cost depends on implementation and its effects on investment incentives.

## 12. Application stores and control over software distribution

Application stores can provide security screening, payment settlement, software discovery and access to large customer bases. They also allow operators to determine participation rules, approval processes and the fees developers pay. A developer whose business depends heavily on one mobile ecosystem may have limited practical alternatives even when competition exists elsewhere in the technology industry.

Restrictions on outside payments, linking or alternative distribution can shape whether an app store merely supplies services or controls access in ways competitors cannot easily contest. These questions have motivated regulatory disputes and court proceedings in multiple jurisdictions. The European Union's Digital Markets Act provides one regulatory approach to designated gatekeepers and their obligations. [6]

A technologically integrated ecosystem can produce legitimate quality and security benefits. A sound analysis does not ignore those benefits, but asks whether particular restrictions are proportionate to them. This is a concrete test of private rulemaking: who can set the rules, which participants can challenge them and how the costs of compliance are distributed.

## 13. Advertising markets and the commercialization of attention

Digital advertising connects the allocation of attention to detailed user information, search behavior and commercial bidding. Platforms can sell targeted access to audiences while mediating what content consumers see. Their revenue may depend on user engagement, and recommendation systems can influence how time is spent. These mechanisms create incentives that sometimes diverge from the interests of users or society.

The concentration of advertising data, distribution and measurement can confer informational advantages. Advertisers may have difficulty independently verifying the full value of impressions, and publishers can depend on a small number of systems for access to customers. Market power may arise through network scale, integrated services and the ability to control performance information.

Calling this technofeudalism is only one possible interpretation. Conventional economics can frame it as a combination of two-sided markets, information asymmetry and monopoly power. The societal stakes nonetheless extend beyond prices to public debate, privacy and the allocation of collective attention. Platform governance becomes a political question when privately designed ranking mechanisms shape which information circulates widely.

## 14. Artificial intelligence as a reinforcement or disruption of platform power

AI could strengthen dominant digital firms because frontier development demands computing capacity, specialized talent and access to data and distribution. The OECD reported that AI absorbed 61% of tracked worldwide venture capital in 2025, up from 30% in 2022. This concentration of financing raises questions about who will own the next generation of information infrastructure. [7]

Large platforms may integrate AI into existing search, advertising, productivity and cloud offerings, benefiting from established customer relationships. Specialized providers may depend on their computing services. Under this scenario, AI amplifies switching costs and permits new forms of information control.

The opposite outcome is also possible. Lower model costs, open-weight systems, new chip suppliers and standardized interfaces can make it easier to compete with established services. AI may reduce the cost of developing applications and migrating work between platforms. Its effect on market structure will depend on actual technical performance, interoperability, regulation and the economics of operating systems at scale. Capital concentration is evidence about financing, not a deterministic forecast of market dominance.

## 15. States and the protection of digital property

Digital infrastructures operate within political jurisdictions. States define enforceable contracts, intellectual-property rights, labor classifications, privacy obligations and tax treatment. They can require platforms to remove certain content, disclose information or restrict access. The platform's authority is therefore substantial but not sovereign in the same manner as an independent state.

At the same time, governments increasingly depend on privately operated cloud, communication and payment services. Procurement can create practical dependence even where formal governmental authority remains intact. Critical public functions may rely on proprietary software and international suppliers, complicating continuity and oversight.

This reciprocal dependence helps explain why the relation between state and platform cannot be reduced to total domination in either direction. Political authority, corporate ownership and technical capability interact. The appropriate institutional analysis asks which actor controls each decision, what alternatives exist and how citizens can challenge outcomes.

## 16. Centre-periphery relations in the digital economy

Digital services cross borders more easily than many physical goods, but the underlying productive capacity remains geographically concentrated. Advanced chips, large data centres, specialized research groups and leading software firms are unevenly distributed. The United Nations Conference on Trade and Development has documented structural asymmetries in the global digital economy, reinforcing the significance of capabilities beyond simple user adoption. [8]

Countries may become important consumers of cloud and platform services while retaining limited control over crucial infrastructure. Revenues, royalties and data-intensive functions can accrue outside the place where final customers live. Local businesses nevertheless may gain from access to wider markets and improved services, creating simultaneous benefits and dependency.

Digital sovereignty is not a binary property. It depends on substitutable providers, workforce capability, standards, cybersecurity, public procurement and the availability of domestic infrastructure. Protectionist measures that raise cost without increasing productive capability may deepen rather than reduce dependence. A stronger development strategy distinguishes the functions essential to national autonomy from services where competitive imports are valuable.

## 17. Brazil: digital payments, commerce and public infrastructure

Brazil combines widespread adoption of digital consumer services with important public financial infrastructure. The PIX payment system illustrates that states and public institutions can establish shared digital rails that support private business innovation. It provides an example of how open access to payments can be structured differently from exclusive platform-controlled payment ecosystems.

Brazilian firms still depend on imported hardware, international software suppliers and large cloud operators for many activities. E-commerce and mobile services can expand market reach while increasing exposure to marketplace fees, advertising costs and platform rules. Smaller companies may have less bargaining power than major national or multinational sellers.

The development question is therefore broader than building a domestic social network or cloud provider. It includes the legal and technical capacity to interoperate, protect personal information, maintain resilient public systems and foster competition. Industrial policies should be evaluated by whether they create sustainable engineering knowledge and access rather than merely subsidize firms displaying a national brand.

## 18. Innovation, concentration and the Schumpeterian counterargument

Joseph Schumpeter emphasized innovation and creative destruction as central to capitalist competition. From this perspective, temporary market power may reward risky investment, while new technologies eventually challenge incumbents. Digital leaders frequently spend heavily on research, acquire equipment and introduce services that lower costs for users. Such activity complicates a purely rent-extractive description.

The strongest counterargument is that profitable platform firms may need scale and data to maintain advanced services. Breaking integration indiscriminately could damage security, interoperability or innovation incentives. Antitrust interventions should therefore evaluate the specific conduct at issue and the benefits that might be lost.

Yet dynamic competition does not guarantee eventual entry. Network effects, acquisitions of potential rivals and technical restrictions can prevent creative destruction. A contestable market requires plausible routes for competitors to reach customers. The Schumpeterian hypothesis must therefore be tested against observed entry, product improvement, pricing and the treatment of firms that challenge incumbents.

## 19. Law, intellectual property and control of infrastructure

Ownership of patents, copyrighted material, proprietary interfaces and infrastructure can create legally protected exclusivity. Those rights may reward research and maintain quality, but their scope and duration influence competition. Digital technologies are especially dependent on interoperability standards and cumulative innovation, so legal restrictions can affect an entire ecosystem of downstream firms.

A company can hold a defensible right to its software while exercising market power over participants who have invested heavily in complementary applications. The economic question is where legitimate exclusion ends and unjustified restriction begins. Competition law, contract law and intellectual-property rules therefore interact rather than operate in isolated spheres.

The technofeudal analogy becomes more precise when used to examine institutional enclosure rather than to claim that private property is intrinsically medieval. The evidence must concern enforceable restrictions, switching costs, the concentration of economic alternatives and the distribution of revenues associated with those restrictions.

## 20. Financial capital, monetary policy and digital platform valuation

Technology platforms operate in capital markets and are influenced by discount rates, credit conditions and expectations of future revenue. Low financing costs can support expansion and acquisitions, while high rates raise the hurdle for investments whose payoff is distant. Their valuations combine expected growth, margins, competitive persistence and market sentiment.

Some platform income resembles a durable annuity when customers face high switching costs. Investors may value such predictability, which can finance additional expansion. This creates a feedback loop: established platforms have more capital to improve services or acquire adjacent capabilities, potentially making entry harder. Yet failed investments and changing preferences can also erode incumbents' positions.

The existence of financial valuation and profit-seeking is an argument against a complete departure from capitalism. At the same time, financial markets may capitalize expected monopoly rents, increasing incentives to defend them through contractual restrictions and lobbying. A coherent political economy examines the interaction between platform property and conventional finance rather than treating them as mutually exclusive systems.

## 21. Taxation and the geographical allocation of digital profits

Multinational digital firms can serve consumers across many jurisdictions with relatively limited physical presence. This complicates the allocation of taxable profits and the measurement of economic activity. Intellectual property, licensing and intragroup transactions can influence where income is recognized. Governments seek to protect tax bases without creating rules that simply transfer costs to local businesses.

Taxes on digital services, corporate profit reforms and withholding arrangements each have different incentives and possible trade consequences. A policy aimed at redistributing some platform rents may also influence service prices, investment and international negotiations. Taxation cannot substitute for all competition policy because a heavily taxed monopoly may remain a monopoly.

Distributional evaluation should consider who bears the burden. A tax formally paid by a platform could be partially passed to advertisers, merchants, workers or consumers. Evidence about pricing responses and contractual terms is therefore more informative than statutory tax incidence alone. Public revenues may support digital infrastructure, education and stronger oversight when governance permits.

## 22. Democracy, public discourse and private moderation

Large information platforms shape visibility through recommendation, ranking, advertising systems and moderation rules. These mechanisms can affect political debate, civil society and the distribution of attention. They create tensions between freedom of expression, protection from abuse, privacy and the need to prevent manipulation.

A private service has legitimate reasons to moderate harmful activity and establish rules, yet concentrated control of major communication channels raises accountability concerns. Users may have limited ability to challenge opaque decisions or understand why certain content is amplified. Laws intended to improve transparency must also guard against the state coercing platforms into suppressing lawful expression.

This political problem is only partly captured by the concept of economic rent. Even if the platform charged no access fee, control over information could be socially consequential. The relationship between ownership, democratic contestation and technological infrastructure requires separate institutional analysis.

## 23. Regulatory responses and their empirical tests

The European Union's Digital Markets Act seeks to establish obligations for designated gatekeepers. Competition authorities have also scrutinized cloud markets, software distribution and digital advertising. Different interventions include interoperability, data portability, bans on selected self-preferencing behavior, contract transparency and structural remedies. [4][6]

A remedy is successful when it changes real market behavior: reduces unjustified switching costs, enables entry, improves consumer options or curtails practices shown to harm competition. It is not enough for a company to announce formal compliance while maintaining equivalent constraints through other rules. Monitoring must consider fees, migration experience, service quality and security.

There are also trade-offs. Mandating technical interoperability can introduce implementation costs or security risks if designed poorly. Structural separation may prevent discrimination but eliminate some operational efficiencies. Institutional design needs case-specific evidence and independent enforcement capacity rather than blanket claims that any regulation either destroys innovation or restores competitive markets.

## 24. What would validate the strong technofeudal hypothesis?

Evidence favoring a strong structural-break claim would show persistent expansion of economic activity that cannot be undertaken outside privately controlled digital territories, accompanied by durable dependence and the ability to impose rules without meaningful alternatives. One would expect ownership of those infrastructures to confer recurring economic claims independent of proportional contribution to production.

The evidence should be comparative. How much income is attributable to control over distribution rather than to services supplied? How quickly do businesses recover when they leave a platform? Are critical functions interoperable? Does the platform determine participants' economic opportunities more strongly than industry regulation or conventional contractual competition? Answers may differ for advertising, cloud, retail and labor platforms.

Evidence against the strongest interpretation includes ongoing competition, significant independent production, entry by new firms, growth of public digital infrastructure and effective constraints imposed by courts and regulators. None of this proves an absence of monopoly power. It helps distinguish a new historical mode of production from a highly concentrated phase of capitalism.

## 25. Three trajectories for platform ownership

In a concentrated-platform trajectory, AI investment, proprietary cloud services and control of consumer interfaces reinforce leading companies. Dependent firms continue paying substantial access fees, and switching remains costly. Digital rent becomes more important to the distribution of income and investment, even though broader capitalist production continues.

In an interoperable-competition trajectory, portability and open standards allow users and developers to move between providers with lower costs. Public digital infrastructure and targeted regulation constrain the unilateral power of gatekeepers. Large firms may remain important but face more credible alternatives and stronger limits on contract terms.

In a fragmented-sovereignty trajectory, states seek national data, cloud and software capacity through regulation, procurement and industrial policy. Security may improve in some functions, while fragmentation raises costs and reduces international interoperability. These paths can coexist across jurisdictions and sectors and do not justify precise probabilistic forecasts without substantial evidence.

## 26. Brazil, development and practical autonomy

For Brazil, the decisive question is not whether to accept the term technofeudalism but how to improve productive bargaining power in digital systems. Firms need competitive access to customers, dependable payments and affordable computing. Workers need clear rights when software controls assignment, remuneration or exclusion. Public institutions need resilient systems and the ability to change suppliers.

Training engineers, maintaining open procurement standards and investing in secure public infrastructure can strengthen autonomy without demanding full national self-sufficiency. Competition policy can target unjustified exclusion, while privacy rules can protect individuals from uses of data they cannot meaningfully contest. Industrial capacity and institutional capability reinforce one another only when policies are evaluated against measurable outcomes.

Imported technology can be useful, and domestic firms can also exercise monopoly power. The objective is not national ownership as an end in itself. It is to preserve access, legal accountability, competitive alternatives and the capacity to innovate without becoming permanently dependent on a single private intermediary.

## 27. A powerful hypothesis requiring historical proof

Technofeudalism offers a provocative account of the relationship between digital ownership, behavioral influence and recurring payments for access. It describes a real transformation in the scale of private infrastructures governing economic exchange. The conceptual dispute remains open because existing capitalism already contains monopoly, rent, financial concentration and coercive dependencies.

Varoufakis identifies novel mechanisms of cloud-based command, Morozov emphasizes continuity with capitalist accumulation, and Durand draws attention to property-driven rentier dynamics. The positions illuminate different aspects of the same industrial transformation. Their value rests on whether they can explain observed prices, contracts, ownership rights and political arrangements better than alternative theories.

The decisive evidence comes from market power, switching costs, the distribution of revenues, labor conditions and the ability of society to govern infrastructure. Whatever name is assigned to the resulting order, democratic institutions face a tangible problem when indispensable commercial and informational systems are privately operated with limited contestability.

## Matrix of mechanisms and evidence

| Mechanism | Possible effect | Evidence needed | Alternative interpretation |
|---|---|---|---|
| Platform commissions | Revenue from access to customers | Costs, margins and alternative channels | Efficient paid intermediation |
| Cloud exit costs | Contractual dependence | Egress fees, migration costs and substitutes | Genuine technical transition costs |
| Algorithmic management | Asymmetric worker control | Decisions, appeal rights and net income | More efficient task coordination |
| Distribution control | Barriers to entry | Data, interoperability and ranking | Legitimate network efficiencies |
| Large AI financing rounds | Concentration of capability | Use of funds and actual customer demand | Productive economies of scale |

```chart
type: bar
title: AI share of worldwide venture investment according to the OECD
unit: % of global financing
2022 | 30
2025 | 61
```

```flow
Network effects and proprietary infrastructure → high switching costs → dependent merchants and developers → capacity to set fees and access rules → platform revenues and margins
Interoperability, portability and competition → lower switching costs → viable alternatives → reduced unilateral market power
```

```mindmap
Platform political economy
- Production
  - Labor and infrastructure
  - Innovation and investment
- Distribution
  - Visibility algorithms
  - Digital marketplaces
- Rents
  - Access commissions
  - Advertising and information
  - Intellectual property
- States
  - Competition policy and taxation
  - Data protection
  - Labor rights
```



```chart
type: bar
title: Composition of global venture funding in 2025, OECD
unit: % of VC
AI | 61
Other sectors | 39
```

```map
title: Institutional geography of platform power
United States | Headquarters of major cloud and platform providers | Corporate headquarters and served markets are not equivalent
European Union | Digital Markets Act and competition enforcement | Measures and duties vary by gatekeeper
Brazil | Public payment infrastructure and digital regulation | PIX does not eliminate dependence on private infrastructure
Developing economies | Dependence on imported digital capabilities | Conditions vary by sector and domestic productive capacity
```

## References
[1] Yanis Varoufakis, *Technofeudalism: What Killed Capitalism* (2023), conceptual discussion: https://www.yanisvaroufakis.eu/2024/01/13/qa-on-technofeudalism-interviewed-by-lorenzo-tecleme-for-valori-it/

[2] Evgeny Morozov, *Critique of Techno-Feudal Reason*, *New Left Review*, 2022: https://newleftreview.org/issues/ii133/articles/evgeny-morozov-critique-of-techno-feudal-reason

[3] Cédric Durand, *Scouting Capital's Frontiers*, *New Left Review*, 2022: https://newleftreview.org/issues/ii136/articles/cedric-durand-scouting-capital-s-frontiers.pdf

[4] UK Competition and Markets Authority, cloud services market investigation, July 2025: https://www.gov.uk/cma-cases/cloud-services-market-investigation

[5] Biglaiser, Crémer, de Cornière and Mantovani, *Dynamic Competition in the Cloud: The Regulation of Egress Fees*, Toulouse School of Economics, June 2026: https://www.tse-fr.eu/publications/working-paper/dynamic-competition-cloud-regulation-egress-fees

[6] Council of the European Union, *Digital Markets Act*: https://www.consilium.europa.eu/en/policies/digital-markets-act/

[7] OECD, *Venture capital investments in artificial intelligence through 2025*, February 2026: https://www.oecd.org/en/publications/venture-capital-investments-in-artificial-intelligence-through-2025_a13752f5-en/full-report.html

[8] UNCTAD, *Facts & figures: The digital economy*, 2025: https://unctad.org/press-material/facts-figures-digital-economy

[9] Nicholas Gane, *Capitalism is capitalism, not technofeudalism*, 2025: https://doi.org/10.1177/1468795X241269293

[10] *Technofeudalism, or capitalism same as it ever was? Placing the blockchain in global capitalism*, *Finance and Society*, Cambridge University Press: https://www.cambridge.org/core/journals/finance-and-society/article/technofeudalism-or-capitalism-same-as-it-ever-was-placing-the-blockchain-in-global-capitalism/3B699270D241F031A2DCCB988DE83E95