Problems with the “Platform Rent Thesis” from the Perspective of Marx’s Theory of Capital Circulation and Their Correction

I. Raising the Question

In today’s world, big data, cloud computing, the Internet of Things, and the new generation of artificial intelligence are accelerating in iteration and being widely applied; the marriage of capital and technology has given birth to and is driving a platform revolution. The platformization of production organization is developing rapidly, and the capitalist factory system is being replaced by the platform system. Among this, digital platforms, as a form of industrial organization different from the traditional industrial economy, have risen abruptly on a global scale, and their influence is growing daily. The platform economy has become an important component of the digital economy, deeply embedded in the digital capitalist economic system and the process of social reproduction. As of September 30, 2025, the total market value of the six leading platform enterprises—NVIDIA, Microsoft, Apple, Google, Amazon, and Meta—was $17.01 trillion. In 2024, the total profits of these super digital platforms reached $476.478 billion, with Google topping the list at $100.118 billion. It goes without saying that digital platforms compete to build or join platform ecosystems, aggregating and retaining various market entities and active users, creating unprecedented digital prosperity through value co-creation.

Whether the丰厚 profits obtained by digital platforms (especially super digital platforms) are essentially “digital surplus value” or “digital ground rent” has sparked extensive academic debate. As things stand, new formulations and viewpoints originating from the “platform rent thesis” are emerging thick and fast: “data colonialism” proposes “data rent”; “techno-feudalism” emphasizes “cloud rent”; “platform capitalism” lists “knowledge monopoly rent,” “advertising rent,” “infrastructure rent,” and so on. In short, from the perspective of platform rent theorists, “the prosperity of the digital economy is fueling a huge rentier economy.” Their logic of argumentation is that digital platforms use modern digital-intelligent technologies to create their own markets and become “gatekeepers,” carrying out digital enclosure to consolidate platform monopolies, ultimately extracting digital ground rent.

Theoretically, the “platform rent thesis” equates digital capital accumulation with the plunder of data resources and the expropriation of surplus value, rarely mentioning the more隐蔽 and exploitative digital labor and its value creation. It thereby magnifies the circulation of digital capital while obscuring the production of digital capital, naturally failing to present the full picture of digital capital circulation under the capitalist mode of production. In practice, the “platform rent thesis” asserts that digital platforms are merely “cloud lords” collecting rent on the spot, and that the monopolistic, parasitic, and predatory nature of digital capital, along with the virtualization of the platform economy, has supposedly become the “dominant” feature, which seems self-evident. Thus, one cannot help but ask: does it not follow that the bigger and stronger digital platforms become, and the more flourishing the platform economy grows, the “worse” digital capitalism will become, even to the point of being unsustainable?

Theory is a reflection of reality. Faced with the cognitive inadequacies and deviations from reality in the “platform rent thesis,” it is urgent to apply Marx’s theory of capital circulation, based on the organic unity of the production process and circulation process of the platform economy, to fully consider the production dimension and value creation of digital platform operations, scientifically reveal the source, form, and essence of digital capital accumulation, and arrive at a new understanding of digital capital circulation that conforms to reality. China is a major digital economy, and its platform economy ranks second globally. Clarifying the functional positioning and production logic of digital platforms, deepening the understanding of the laws governing digital capital movement, and advancing the innovative development of Marx’s theory of capital circulation are of great theoretical and practical significance for grasping the characteristics and behavioral patterns of digital capital within the socialist market economy and promoting high-quality development of China’s platform economy.

II. The Non-Production Narrative of the “Platform Rent Thesis” and Its Deviation from Reality

Entering the digital-intelligent era, various forms of digital platforms are emerging competitively, increasingly exhibiting a composite nature that spans both the circulation and production stages, integrating different forms of capital. Broadly speaking, the “platform rent thesis” remains within the sphere of circulation of digital capitalist social reproduction, asserting that digital platforms are merely “rentiers” of digital assets, relying on platform monopolies to enslave users and exploit third parties. But reality is far from this. In the global wave of digital industrialization and industrial digitalization, the productive investment activities of digital capital, represented by leading platform enterprises, are very active, frequently launching digital products and digital services favored by consumers. The non-production narrative of the “platform rent thesis” clearly diverges from the vibrant productive practices of digital platforms.

1. The Non-Production Narrative of Digital Platforms as “Rentiers”

According to the common definition, a digital platform is a form of business organization that uses network information technology to enable interdependent bilateral or multilateral parties to interact and match under specific rules, thereby co-creating value. Essentially, digital platforms are the real-world carriers of the value movement of digital capital. On this point, platform rent theorists believe that platforms are becoming territories. That is to say, digital platforms, by virtue of algorithmic control, data monopolization, and the privatization of digital infrastructure, have suddenly become the “new landlords” of the digital-intelligent age. Consequently, “the logic of the system of digital capitalism itself is undergoing a qualitative transformation,” where accumulation by dispossession replaces productive accumulation, ground rent takes precedence over profit, and rent replaces profit.

First, digital platforms are keen on “digital enclosure,” wantonly extracting “data rent.”

As the networking, digitalization, and intelligentization of human production and life deepen, data resources are experiencing explosive growth. The process of data factorization and valorization, led by super digital platforms, is accelerating and intensifying. Undeniably, digital capital that firmly controls the platform ecosystem, in acquiring data resources, on the one hand “legally” occupies the data of active users through user service agreements; on the other hand, it “freely” collects global public network data by using digital-intelligent technologies. Based on this, proponents of the “platform rent thesis” argue that the increasingly vast network data resources are coveted by digital capital, and that digital platforms, through “digital enclosure” and “data colonialism,” achieve a “force-of-nature-like” utilization of massive amounts of data. Consequently, “any platform enterprise that forms a monopoly can obtain data rent through data control, which is analogous to absolute ground rent.”

Second, digital platforms seek a monopoly on digital intellectual property, striving to extract “cognitive rent.”

In the current era, the monopoly of digital intellectual property has become a new form of capitalist intellectual property monopoly. In reality, internet technology giants, relying on their雄厚 capital strength, powerful digital-intelligent technological advantages, and pioneering digital industrialization practices, monopolistically possess increasingly vast digital knowledge assets, such as databases, operating systems, intelligent algorithms, large AI models, and digital brands. Furthermore, digital platforms leverage their structural power within the platform ecosystem to meticulously set up a global accumulation system for monetizing digital intellectual property. From the perspective of platform rent theorists, the profits from the knowledge monopoly of digital platforms are essentially an economic distribution mechanism based on predatory appropriation. The reason for this is that the digital knowledge assets held by data platforms do not primarily originate from innovation, but rather depend on the accumulation by dispossession of the human knowledge commons. Whether it is the uncompensated appropriation of the “general intellect” of digital cyberspace or the privatization of gradually accumulated digital-intelligent technological knowledge originally resulting from public R&D, the large-scale enclosure of knowledge by digital platforms means that “now human knowledge is increasingly being enclosed, privatized, or turned into commodities.” The result is that, for digital capital, seeking and consolidating a monopoly on digital intellectual property has become a powerful means of extracting digital surplus value.

Third, digital platforms become the “gatekeepers” of online markets, automatically extracting “transaction rent” and “traffic rent.”

Entering the stage of digital capitalism, digital capital uses intelligent algorithms to realize a "connection revolution," develops the market transaction functions of digital network space, and drives the formation of a multi-sided market within the platform ecosystem. Proponents of the platform rent thesis argue that "platform enterprises do not possess the means of production, but rather create the means of connection." Super digital platforms, as the "gatekeepers" of online markets, firmly control platform access rights, manipulating the allocation and management of online transaction opportunities; at the same time, they dominate public-domain traffic and centralized operations, achieving control and distribution of traffic resources within the platform ecosystem. Consequently, on the one hand, in online market transactions such as Amazon's online marketplace and Apple's app store, digital platforms take a cut per transaction, collecting substantial market settlement fees; on the other hand, in digital marketing such as Google and Meta's online advertising and algorithmic recommendations, digital platforms "monetize" user attention, obtaining a steady stream of "traffic rent." In the context of the "platform rent thesis," digital platforms' collection of online settlement fees and "traffic rent" ultimately relies on digital capital's monopoly over scarce resources in the digital economy.

Fourth, digital platforms deploy digital infrastructure and collect "cloud rent" from a position of dominance.

Entering the 21st century, a notable feature of large-scale production in capitalist society is the infrastructuralization of digital platforms and the platformization of infrastructure. Broadly speaking, the network infrastructure of the Internet of Everything, the computing infrastructure integrating "cloud, network, terminal, edge, and application," and the application infrastructure combining digital-real integration and cross-domain collaboration have become indispensable general conditions of production for the value movement of digital capitalism. For this reason, "whoever controls these infrastructure concentrates political and economic dominance over those who participate in them." In reality, super digital platforms such as Amazon, Microsoft, and Google have seized the initiative through global deployment, controlling the "cloud foundation" of digitalized and intelligent production networks, becoming the primary owners of global digital infrastructure, and then, through digital-intelligent transformation, incorporating more and more market entities into privatized platform ecosystems. In the view of platform rent theorists, super digital platforms control global digital infrastructure, forming and accumulating "cloud capital," algorithmically exploiting "cloud proletarians" and "cloud slaves," while simultaneously collecting "cloud rent" from third parties dependent on the "digital land" of the platform.

2. Digital Platforms ≠ "Rentiers": The Stylized Facts

Global practice in the platform economy shows that digital platforms represented by internet technology giants are by no means detached and idle "rentiers" of digital assets. This is because they themselves directly provide products or services. In response to the deviation of the "platform rent thesis" from reality, the contemporary Western left-wing scholar Evgeny Morozov's question is striking: "If internet giants were truly lazy rentiers, relying on intellectual property and network effects to harvest wealth, why would they invest huge sums in various productive activities?" From the perspective of the production reality of digital capitalism, digital platforms play a key and significant role in the formation of data factors, digital-intelligent technological innovation, digital space production, and the supply of digital-intelligent services. Broadly speaking, the productive nature of digital platforms is manifested in the following four aspects.

First, digital platforms are key producers in the formation of data factors.

Entering the digital-intelligent era, the structure of factors of production is undergoing transformation, showing a trend of "labor-intensive → capital-intensive → technology-intensive → knowledge-intensive → data-intensive." It is indisputable that data has become a new key factor of production in capitalism. In this regard, the "platform rent thesis" simplistically analogizes data to land resources, concluding that once digital platforms predatorily possess massive amounts of data, they can sit back and reap a steady stream of "data rent." However, from the perspective of the full data lifecycle, data is different from data resources, data resources are not equivalent to data factors, and neither can be conflated with data commodities. From the perspective of the data industry chain, whether it is digital footprints or user-generated content in the consumer internet, or production data or intelligent supply chain data in the industrial internet, disorderly and chaotic raw data requires substantial financial, human, and time investment. It must undergo engineering processing such as processing, storage, and management to become high-quality, multi-modal, specialized data resources—data factors—which can then be rapidly integrated into various scenarios such as R&D design, flexible production, intelligent manufacturing, precision marketing, market promotion, and enterprise management, thereby opening up new industries, new business formats, and new models, and realizing a multiplier effect on economic and social development.

In this regard, digital platforms not only collect and store data but also utilize their own technological advantages and application scenario needs to process, analyze, and empower data applications, becoming a central force in data integration and data mining. As one report points out, "As owners and operators of massive data, platform enterprises have a natural advantage in releasing the multiplier effect of data factors and are a powerful entity driving the transformation of data factors into productive forces." It is certain that with the rising demand of the new generation of AI, represented by large AI models, for high-frequency multi-dimensional datasets and high-quality corpora, the productive role of digital platforms in data scraping, data storage, data annotation, data analysis, and data governance will become increasingly prominent.

Second, digital platforms are a central force in modern digital-intelligent technology R&D.

Entering the 21st century, a new generation of digital-intelligent technologies, represented by big data, cloud computing, artificial intelligence, blockchain, mobile internet, the Internet of Things, virtual reality and augmented reality, and quantum computing, have emerged and accelerated in iteration. They increasingly drive the networking, digitalization, and intelligentization of all aspects of human production, distribution, exchange, and consumption, and have become the technological foundation for the value movement of digital capitalism. It goes without saying that these disruptive digital-intelligent technological innovations typically require long R&D cycles and high demands for practical application, necessitating substantial patient capital for long-term support.

In this regard, super digital platforms have significant advantages in funding, talent supply, and innovation ecosystems, naturally becoming the main force in global digital-intelligent technology innovation. The total R&D spending of the five super digital platforms—Meta, Google, Apple, Microsoft, and Amazon—surged from $4.4 billion in 2000 to $199 billion in 2022. Especially in the wave of intelligent development, super digital platforms have increased R&D investment, competing to engage in a new round of digital-intelligent technology competition. In 2024 alone, the average R&D expenditure of seven US platform companies—Apple, Microsoft, Google, Amazon, Meta, Netflix, and Uber—reached $33.26 billion, a year-on-year increase of 12.0%, with an average R&D intensity of 13.5%. In terms of "hard tech" output in the digital economy, super digital platforms have almost dominated the development and application of operating systems, search engines, intelligent chips, application products, and large AI models. Stanford University's "AI Index Report 2025" states that among the notable large models released globally between 2014 and 2024, the proportion from the industry sector steadily increased, reaching 90.2% in 2024; Google led with 186 models, followed by Meta (82) and Microsoft (39). There is no doubt that super digital platforms, possessing big data, massive computing power, and top talent, have become key drivers and frontier leaders in digital-intelligent technology R&D and AI innovation.

Third, digital platforms are the primary implementers of digital capital's spatial production.

“Space is an element of all production and all human activity.” Since the establishment of the capitalist mode of production, capitalism has perpetuated the logic of capital through the occupation of space and the restoration of space. Entering the stage of digital capitalism, digital capital has become an important form of capital movement, and the digital network space, which breaks through traditional spatiotemporal limits, has become the main field for digital capital to absorb labor. In reality, the formation and maintenance of digital network space, on the one hand, relies on the platform access and active participation of active users, and on the other hand, requires substantial capital investment in the basic network architecture of the platform, user interface design, and platform ecosystem maintenance.

In this regard, digital platforms represented by internet technology giants actively deploy digital infrastructure globally, constructing and consolidating the material foundation for the spatial production of digital capital. At the same time, digital platforms actively engage in algorithm development and application scenario innovation, striving to create platform ecosystems with diverse economic and social functions, thereby realizing the capitalization of digital network space. Among them, media platforms such as Google and Meta [n] build new spaces for online marketing through information linking, virtual social interaction, and algorithmic traffic generation; transactional platforms like the Apple App Store and Amazon Online Marketplace use algorithmic matching and automatic recommendations to open up new fields of online markets and digital circulation; lean platforms such as Deliveroo [n] rely on on-demand services and algorithmic management to bridge the “last mile” of the gig economy; industrial cloud platforms like Siemens Industrial Cloud and GE Industrial Cloud achieve comprehensive interconnection of people, machines, and things, constructing collaborative production and digital twin spaces across the entire factor, value chain, and value stream; product platforms like Apple Music focus on personalized customization of digital content, opening up new spaces for contemporary people’s spiritual and cultural consumption. Furthermore, digital platforms also construct algorithmic rules, credit evaluation systems, and information filtering mechanisms to effectively reduce information asymmetry and moral hazard, establish a trustworthy online transaction environment, improve market operational efficiency, and provide a strong institutional guarantee for the efficient matching of supply and demand for goods or services.

Fourth, digital platforms are the main suppliers of digital-intelligent services.

Entering the stage of digital capitalism, digital platforms, based on the production logic of “data + model = service,” leverage the advantages of broad connectivity, efficient matching, and dynamic optimization to achieve the commercial-intelligent transformation of massive data. While providing and continuously optimizing basic platform services for free, they also offer various digital-intelligent services aimed at expanding markets and creating markets for the core transactions within the platform ecosystem. Broadly speaking, from free services like Google Search, Meta’s social networking services, and Instagram’s photo-sharing services, to commercialized services like Uber’s on-demand services and Amazon Mechanical Turk’s crowdsourcing services; from cloud computing services provided by Amazon, Microsoft, and Google to the full-stack development services for large AI models that internet giants are competing to offer, digital platforms play an increasingly important role in all aspects and processes of human social production and daily life.

Digital platforms provide a wide variety of digital-intelligent services, extensively penetrating and being applied in all links of social reproduction, including production, distribution, exchange, and consumption, within digital capitalism. They not only promote the deep development of human digital-intelligent existence, bringing unprecedented digital convenience and digital experiences to the general public, but also, through the integration of digital and real economies, platform empowerment, and intelligent driving, effectively enhance total factor productivity, significantly reduce market transaction costs, generate increasing returns, and achieve digital prosperity.

III. The Theoretical Simplification of Digital Capital Circulation by the “Platform Rent Thesis”

Entering the digital-intelligent era, although the operational methods of each digital platform differ, the primary goal of digital platforms is to “match users, creating value for all participants through the exchange of goods, services, or social currency.” In this regard, a common interpretation of the “platform rent thesis” is that digital platforms are merely “market intermediaries” providing internet infrastructure, facilitating supply-demand matching but not engaging in any actual production. In this context, digital capital is not industrial capital; digital capitalists operate outside the actual production process, and thus the theoretical examination of digital capital circulation can only be confined to the sphere of circulation, which mediates the reproduction process.

1. Theoretical Simplification of Digital Capital Circulation

Marx pointed out that the valorization process of capital is the unity of the production process of capital (value production) and the circulation process of capital (value realization). From a micro perspective, the movement of capital value is manifested as a single industrial capital completing the full cycle of G—W…P…W′—G′, that is, the circulation of capital sequentially passes through the stages of purchasing means of production and labor power, production, and sale, sequentially assuming the three functional forms of money capital, productive capital, and commodity capital, corresponding respectively to the preparation, production, and realization of surplus value. In contrast, as a special form subordinate to industrial capital, the function of commercial capital is “to specifically mediate commodity exchange,” and thus its capital circulation is confined to the sphere of circulation, achieving capital valorization through two or more transfers of commodities, in the movement form of G—W—G′.

The “platform rent thesis” holds that digital platforms no longer invest in products but create “connections.” Taking the Uber platform as an example, “unlike Ford Motor Company, Uber does not own any value available for production and consumption; what it owns is merely the way to connect users—not the way of production.” Furthermore, although the “connection” functions displayed by digital platforms are diverse in form, their fundamental way of making profit is to use massive data for algorithmic driving, promoting precise, efficient, and timely matching of supply and demand in the online market, thereby helping traditional industrial capital shorten circulation time, increase the speed of capital turnover, and indirectly increase the surplus value produced by industrial capital. For this reason, various forms of digital platforms are essentially a new type of commercial capital. From the perspective of mediating the circulation process of industrial capital, digital platforms open up the digital circulation sphere, act as “gatekeepers” of the online market, and thus make huge profits by extracting massive data and precisely placing advertisements. Consequently, “although digital platforms obtain profits by utilizing audience labor, this is the same in nature as the profits obtained by commercial capital.” That is to say, platform organizations need to rely on non-platform organizations to complete the cycle of valorization.

Therefore, in the context of the “platform rent thesis,” the value movement of digital capital does not have the production process as its intermediary, nor does it need to undergo the W—W′ change as traditional commercial capital must. This means that the market intermediary logic of digital platform operations makes the valorization movement process of digital capital more similar to interest-bearing capital: it does not take the forms of productive capital and commodity capital, only the form of money capital, and is always situated in the digital circulation sphere of capital. The result is that, in the basic propositions of the “platform rent thesis,” “everything that happens in the middle disappears” in the movement of digital capital; all the various real elements mentioned earlier—data factor production, digital-intelligent technology research and development, digital infrastructure construction, and digital-intelligent service provision—are written off, meaning that in the case of digital capital, “the movement of capital is simplified; the intermediary process is omitted.” Under the theoretical simplification of the “platform rent thesis,” digital capital circulation is presented in the form of G—G′, as if in the digital-intelligent age capital has acquired an even more automatic and pure “fetish form.”

2. The Root of the Cognitive Deviation Regarding Digital Capital Circulation

There is no doubt that business model innovations on digital platforms are emerging thick and fast today. At the same time, driven by the twin engines of digital industrialization and industrial digitalization, the movement of digital capital is interpenetrating and merging with traditional industrial capital and financial capital, objectively creating a complex situation in which the valorization of digital capital is both productive and predatory, and profit formation coexists with rent extraction. Fundamentally, the “platform rent thesis” misapprehends the circulation of digital capital because it magnifies the non-productive character of digital capital and lacks any basic regard for the productive practices of digital platforms. It must be pointed out, however, that the “platform rent thesis” has a fundamental defect in its understanding of the movement of digital capital; it is not a deliberate distortion of digital platforms and the platform economy. As Evgeny Morozov observed when questioning the “techno-feudalism” current of thought, “innovations of perspective” of this sort are in fact incapable of grasping the production logic of the digital economy; they cannot figure out what digital platforms actually produce and how they produce it. The platform-rent theorists’ one-sided understanding of the functional positioning of digital platforms and the characteristics of digital capital stems from several factors.

First, the prevalence of the free economy creates an “optical illusion” that obscures the productive dimension of digital platforms. In the digital capitalist stage, leading platform enterprises efficiently develop or provide a variety of digital-intelligent services. Since the deployment of large AI models in production, “one-click generation” has created an unprecedented digital abundance, and most of these digital-intelligent services are provided free of charge. As Chris Anderson put it, “In the digital marketplace, free is always a choice available to consumers.” For example, Google, a global internet giant, offers nearly a hundred free services — from search engines and academic literature retrieval to photo-editing software, word processors and spreadsheets, from YouTube video sharing to the AI tool Gemini that Google has launched. These are genuinely free, not some kind of hard-sell tactic. It goes without saying that the vast array of digital-intelligent products and platform services constitutes a massive digital wealth and produces significant economic and social effects.

Undoubtedly, the current business models of digital platforms are emerging in an endless stream. At the same time, driven by the dual forces of digital industrialization and industrial digitalization, the movement of digital capital is interpenetrating and integrating with traditional industrial capital and financial capital. This objectively creates a complex situation where the valorization of digital capital exhibits both productive and predatory characteristics, and profit formation coexists with rent extraction. Fundamentally, the “platform rent thesis” misapprehends the circulation of digital capital because it exaggerates the non-productive nature of digital capital and lacks a basic consideration of the productive practices of digital platforms. However, it must be pointed out that the “platform rent thesis” has a fundamental flaw in its understanding of the movement of digital capital. This is not a deliberate distortion of digital platforms and the platform economy. As Evgeny Morozov, questioning the trend of “techno-feudalism,” pointed out, such “innovations in perspective” are actually incapable of discerning the production logic of the digital economy and fail to understand what digital platforms are producing and how they produce. The biased understanding of the functional positioning of digital platforms and the characteristics of digital capital among platform rent theorists stems from several factors.

First, the prevalence of the free economy creates an “optical illusion” that obscures the productive dimension of digital platforms. Entering the stage of digital capitalism, leading platform enterprises, such as major platform companies, efficiently develop or provide various digital-intelligent services. After the implementation of large-scale AI models, “one-click generation” has created an unprecedented digital boom, and most of these digital-intelligent services are provided for free. As Chris Anderson stated, “In the digital marketplace, free is always a choice available to consumers.” For example, as a global internet giant, Google provides nearly a hundred free services, ranging from search engines and academic literature retrieval to photo editing software, word processors, and spreadsheets, from YouTube video sharing to Google’s AI tool Gemini. These are genuinely free, not some kind of aggressive marketing tactic. Needless to say, the vast array of digital-intelligent products and platform services forms a massive digital wealth, generating significant economic and social effects.

As is well known, under the capitalist mode of production, the capitalist “must produce not only use-value but also a commodity, not only use-value but also value, not only value but also surplus value.” From the perspective of the form of capitalist wealth, “to become a commodity, a product must be transferred by exchange to the person for whom it serves as a use-value.” In this regard, the numerous digital-intelligent services provided for free by digital platforms do not undergo market exchange mediated by money. They possess social use-value but do not constitute value. From the perspective of platform rent theorists, platform users, as consumers, enjoy free services, while digital platforms merely harvest data and user attention, failing to engage in direct value production. Thus, the non-productive nature of digital platforms and the “predatory” nature of digital capital accumulation are considered self-evident.

Second, the self-promotion of digital capital as a “middleman” reinforces the non-productive appearance of digital platforms. In practice, digital platforms often describe themselves as “intermediaries” or “markets,” whose role is limited to providing matching transaction services. From a political economy perspective, digital platforms deliberately “self-deprecate.” On the one hand, this is a “shrewd move” to cater to capital markets, meaning that “to keep the corporate prestige, financial networks, and technological landscape of the platform intact, workers must be kept out of sight.” On the other hand, it enables “regulatory arbitrage” in legal gray areas, striving to completely shed the employer obligations of digital labor. This self-marketing rhetoric of digital capital subtly influences the social perception of how digital platforms actually operate.

Third, Marx’s assertion that the circulation of capital realizes value is mechanically applied to the actual understanding of the value movement of digital capital. Marx pointed out, “The circulation of capital realizes value, just as living labor creates value.” Accordingly, a basic view of Marx’s theory of capital circulation is that “the circulation process is a phase of the total reproduction process. But in the circulation process, no value is produced, and therefore no surplus value is produced.” For this reason, capital performing the function of commodity exchange is non-productive, and this does not change even if the function of commodity capital becomes independent. Consequently, the labor of workers engaged in the sphere of commodity circulation under capitalism is merely “labor that performs an intermediary function” and “does not directly produce surplus value.” In short, “merchant capital does not directly participate in the creation of profit or surplus value; it obtains its share from the quantity of profit produced by industrial capital, in proportion to its share in the total capital.” From the perspective of platform rent theorists, since the business activities of digital platforms are confined to the sphere of digital circulation, they are essentially “mere circulation agents.” Therefore, their theoretical characterization of the source of digital capital valorization as a transfer of profit from industrial capital seems reasonable. In this regard, the view of a recent study is quite pertinent: if the dichotomy of production-circulation and productive-nonproductive is applied too rigidly to the commodification process in the digital-intelligent age, it is highly likely to overlook the fact that capital is using digital-intelligent technology to produce value in unprecedented ways.

3. The Harm of Simplifying the Theory of Digital Capital Circulation

Entering the 21st century, the development of the digital economy is flourishing, and various platform services are increasingly becoming basic elements of human production and daily life. Theoretically, the identification of the actual source of profit for digital platforms directly relates to the definition of the nature of the platform economy and the grasp of the characteristics of digital capital. In this regard, as discussed above, the “platform rent thesis” claims that digital platforms merely “reap without sowing,” and even asserts that this new mode of production is reactionary. Clearly, allowing such views to spread and grow will inevitably bring about various practical harms.

Third, platform regulation faces cognitive misdirection. The practice of the platform economy shows that digital capital both displays a positive role in developing digital productive forces and expanding the boundaries of human production possibilities, and increasingly reveals a destructive side, including platform monopoly, algorithmic exploitation, and digital colonization. Only by accurately grasping the characteristics and behavioral laws of digital capital can government platform regulation avoid a policy “pendulum” — that is, merely relying on platform self-discipline, or depending on strict regulation as a “once-and-for-all” solution. In this regard, the “platform rent thesis” magnifies the circulation intermediary function of digital platforms, overstates the predatory nature of digital capital and the virtualization tendency of the platform economy, and inevitably creates cognitive interference for the rational design and policy choices of government departments in platform regulation.

IV. The Commodification of Digital-Intelligent Services and the Theoretical Correction of Digital Capital Circulation

In the stage of digital capitalism, the operational activities of digital platforms have become increasingly complex and diverse. The value movement of digital capital has transcended the unidirectional linear value chain and moved toward a platform value matrix that integrates digital and real economies and aggregates networks. To fundamentally correct the theoretical simplification of the value augmentation process of digital capital by the “platform rent thesis,” it is necessary to explore the production secrets of the digital platform as a “digital factory” and the “value growth domain” of the platform economy. Then, in the practice of commodifying digital-intelligent services, the true face of digital capital circulation can be dialectically grasped and scientifically presented.

1. The Commodification of Digital-Intelligent Services

As is well known, the commodity is the “elemental form” of capitalist wealth. Entering the digital-intelligence era, data has become a new key factor of production, and the production and trading of digital products based on data collection, processing, and analysis constitute a new picture of value movement in contemporary capitalism. Looking at the typical forms of digital products, “unlike the era of industrial capitalism, the key to profit-making in digital capitalism is not mass-producing uniform products, but providing personalized, customized services for different customers.” On this point, even the platform rent theorists admit that user-generated data enables digital platforms “to improve user experience, design targeted advertisements, or sell personalized services.” For this reason, by clearing away the practical fog of free platform operations, clarifying the practical patterns of their commodification of digital-intelligent services, and penetrating the production logic of digital capital movement, it is no longer difficult to theoretically grasp the production process of digital capital circulation.

First, data-driven and business intelligence services. In the digital capitalist stage, with the rapid development of a new generation of artificial intelligence technology represented by machine deep learning, the value of data as “fuel” for intelligent machines continues to be released. Broadly speaking, “no data, no prediction.” In the application of analytical AI technologies, data-driven predictive analysis technologies are developing rapidly. Digital platforms leverage the synergistic advantages of economies of scale and scope, applying the resulting digital-intelligent solutions widely in fields such as inventory management, credit rating, facial recognition, fraud risk detection, health insurance, and social governance. In recent years, with the everyday application of generative AI technologies, repetitive and rule-based knowledge work tasks — such as writing, translation, tabulation, charting, and code checking — are efficiently completed automatically through “one-click generation” by large AI models.

In practice, AI content generation (AIGC), with its rapid response capability, vivid knowledge output, and rich application scenarios, plays an increasingly important role in accelerating the intelligent transformation of human production and daily life. In particular, as the digital economy and the real economy deeply integrate, leading platform enterprises use industry knowledge to build vertical large models or scenario-specific large models, providing generative business intelligence services on a large scale, including digital twins, industrial design, drug simulation, power grid modeling, video generation, and animation rendering. Whether for consumer services (ToC) or enterprise services (ToB), “no data, no decision” and “no data, no generation” are driving and promoting a cognitive revolution in digital capitalism. Furthermore, under emerging business models such as “model as a service,” “inference as a service,” and “agent as a service,” digital capital charges fees based on model customization or per-token input and output, thereby completing the value loop of business intelligence services.

Second, algorithmic matching and precision marketing services. In the current era, the application of digital-intelligent technologies has revolutionarily reshaped the sphere of capital circulation, greatly dissolving the spatial and temporal constraints on value realization. Looking at platform economy practices, digital platforms, on the one hand, aggregate multiple market actors — including advertisers, service providers, producers, and consumers — to develop long-tail markets, reduce search costs, match supply and demand, and create online transaction opportunities. On the other hand, they open up virtual spaces for interaction, create various virtual communities, frequently generate online hot topics, and develop a traffic economy. Within this, digital platforms analyze massive data, identify user preferences, form algorithmic recommendations, and automatically push “thousands of faces for thousands of people” or even “one thousand faces for one person” precision advertisements, greatly improving circulation efficiency, helping traditional producers reduce their “individual turnover time,” and indirectly expanding the production of surplus value. For this reason, media platforms that rely on digital advertising for profit, such as Google and Meta, actually sell a special commodity — “promotional services.” As some scholars point out, “equating advertising revenue with the value of Google’s search service output would be a misunderstanding.” On this point, the “platform rent thesis,” confined to the traditional understanding that marketing activities in the circulation sphere are non-productive, and mistakenly treating numerous active users as actual consumers, naturally fails to perceive the production logic of digital marketing services, and any realistic examination of digital platforms as direct producers becomes impossible.

Third, intelligent production and cloud computing services. As things stand today, the networking, digitalization, and intelligent transformation of human production and daily life are accelerating, and the actual demand for cloud adoption and usage is experiencing explosive growth. In particular, with the application of a new generation of AI represented by large models, multi-source heterogeneous data scales are expanding, algorithm models are becoming increasingly complex, and the requirements for computing power and parameter tuning speed are rising rapidly. Cloud computing, relying on its own powerful elastic computing capacity and massive data storage capability, can quickly meet the efficiency requirements of AI infrastructure for various clients — including enterprises, research institutions, and government departments — in their efforts to “go to the cloud, use data, and empower intelligence.”

In this regard, cloud service providers represented by Amazon, Microsoft, and Google effectively integrate resources, achieve componentized and modular encapsulation of digital infrastructure capabilities, and provide efficient, low-cost integrated services to meet the needs of diverse scenarios in industrial manufacturing, finance, education, healthcare, and public services. Data shows that in 2024, the global cloud computing market size, represented by “infrastructure as a service,” “platform as a service,” and “software as a service,” was USD 692.9 billion, with a year-on-year growth rate of 20.3%. It is projected that by 2030, the global cloud computing market size will approach USD 2 trillion. In 2024, the cloud computing business revenues of Amazon, Microsoft, and Google reached USD 107.6 billion, USD 102.4 billion, and USD 43.3 billion, respectively. For small and medium-sized enterprises, the existence of the cloud computing market transforms what were originally high fixed capital expenditures into controllable digital-intelligent operating costs, thereby effectively resolving their own constraints in funding, talent, and management, enabling them to go to the cloud with light assets and use the cloud at low cost, quickly and conveniently completing intelligent transformation and digital transformation, and accelerating the formation of intelligent productive forces. On this point, even the platform rent theorists recognize that in reality, many enterprises do not build their own digital infrastructure but instead choose to purchase cloud computing services from super digital platforms.

2. The Value Creation of Digital Labor

Entering the stage of digital capitalism, the various forms of digital-intelligent services that are increasingly automated and “unmanned” are in fact inseparable from the online input and key support of digital labor. At present, the digital labor force can be broadly divided into two categories: one consists of digital-intellectual technology elites formally employed by digital platforms, who are the “coding experts” of the platform economy and the “labor aristocracy” of the digital age; the other comprises freelancers, crowdsourced micro-workers, and gig workers from ride-hailing platforms, forming the “cyber-proletariat.” These digital laborers are numerous, but most are trapped in an “algorithmic cage” and face precarious living conditions. In practice, although digital capital deliberately obscures the value contribution of digital laborers through the concealment of labor, a production truth gradually emerging is that the widespread reification of digital labor constitutes an important source of profit formation for digital platforms. In response, the “platform rent thesis” favors a narrative of digital labor exploitation as “serfdom” or “servitude,” denying or deliberately avoiding the irreplaceable role of digital labor in the supply of data factors, the production of digital space, the manufacture of intelligent machines, and the production and sale of digital commodities.

First, from the perspective of factor input in digital-intelligent services, the quantitative growth and qualitative improvement of data factors are the collective product of digital labor. In reality, hundreds of millions of active users compete to engage in digital content publication, webpage creation, likes, reposts, ratings, and comments, and the user-generated content thus produced freely contributes an endless stream of data “raw materials” for the intelligent production of the “digital factory.” Especially after the deployment of large-scale AI model production, the quality and richness of user-generated content have continuously improved under human-machine collaboration, and the data value of users’ unpaid labor has become increasingly evident. Similarly, gig workers rushing about on the “last mile of artificial intelligence” constantly generate vast amounts of data that serve the optimization of platform algorithms, often without their awareness. Furthermore, large quantities of unstructured data, full of “noise,” are identified and labeled by crowdsourced micro-workers around the world, achieving the productive transformation of data resources into data factors, to the point that “the magic of machine learning all lies in the arduous labor of data annotation.”

Second, from the perspective of the means of production supporting digital-intelligent services, the intelligent machine system, as “objectified knowledge power,” is an innovative creation of digital labor. Entering the stage of digital capitalism, a notable sign of the development of social productive forces is the emergence of intelligent machines in the process of the dematerialization of fixed capital. The intelligent machine system, represented by platform algorithms, deeply processes vast amounts of data to form digital-intelligent services with various functions, bringing unprecedented convenience in digital life and prosperity in digital wealth. For example, digital capital hires top talent in digital-intelligent technology at high salaries, focusing on core algorithm research and development, platform architecture design, and general intelligent agent development—highly creative forms of advanced digital labor—continuously enhancing the “human-like intelligence” of the digital capitalist machine system, thereby accelerating the iteration of digital-intelligent service supply and driving innovation in digital platform profit models. It should be noted that in the increasingly “fully automated” production of digital capitalism, the value contribution of employed digital laborers, such as data scientists, machine learning engineers, AI quality controllers, and AI ethics consultants, is largely realized through value transfer in the “one-click generation” of intelligent machines.

Finally, from the perspective of the production-consumption闭环 of digital-intelligent services, the high-quality supply of platform services is the result of human-machine collaboration in digital labor. Entering the digital-intelligent era, the anthropomorphism and agency of intelligent machine systems are increasingly strengthened, and the human-machine relationship in digital capitalism is rapidly moving from “human-machine adaptation” to “human-machine collaboration,” forming a “hybrid intelligence” that jointly drives the improvement of quality and efficiency in digital-intelligent services. From human scoring in Google’s search engine to content moderation in metaverse social links, from problem design, result verification, and complex decision support in analytical AI applications to reinforcement learning and value alignment in large-scale AI model development, digital labor always calibrates quality and “turns stone into gold” in the production and consumption of digital-intelligent services through “human-machine interaction.” Moreover, in the march toward intelligent survival, every close encounter between active users and general intelligent agents—whether asking questions and chatting, or providing feedback and corrections—contributes, without compensation, to improving the quality and efficiency of platform digital-intelligent services.

3. Theoretical Presentation of Digital Capital Circulation

Fundamentally, digital capital is a capital complex with complex attributes and variable characteristics. From the perspective of its mediating role in the circulation of traditional industrial capital, digital capital is commercial capital active in the digital circulation sphere; from the perspective of the commodification of digital-intelligent services, digital capital is a new type of industrial capital that has fully permeated the economic system alongside digital-intelligent transformation; from the perspective of the predatory trend of capital accumulation, digital capital also reveals characteristics and behavioral patterns similar to those of fictitious capital. For this reason, only by applying the basic principles of Marx’s theory of capital circulation to comprehensively present the actual process of digital capital circulation can we deepen our understanding of the characteristics and behavioral laws of digital capital and scientifically reveal the true picture of value movement under digital capitalism.

First, the capital circulation of digital capital as a new type of industrial capital.

Under the capitalist mode of production, the organic unity of the production process and the circulation process, and of value creation and value realization, constitutes the basic content of industrial capital circulation. Theoretically, any money capital, whether invested in the production sphere or the circulation sphere, as long as it employs labor to form products or services and realizes capital增殖 through market transactions, then possesses the basic attributes of industrial capital. Entering the digital-intelligent era, as discussed above, whether it is a circulation-oriented digital platform or a production-oriented digital platform, as long as it provides digital-intelligent services in platform operations and commercializes them, it in fact creates new value and realizes the production of surplus value.

In detail, at the micro level of capital movement, like traditional industrial capital, digital capital also forms a complete cycle of G—W…P…W′—G′. That is, in the purchase phase, digital capital procures hardware and software systems for intelligent production, builds or rents data centers, and employs advanced digital laborers; in the production phase, platform laborers process and handle vast amounts of data, develop and iterate algorithms, and form digital-intelligent products and services with various functions; in the sales phase, digital platforms innovate business models, achieving commercial profitability through subscription fees, advertising fees, model customization, and other means. At the meso level of industrial development, the industrialization of digital technology and the digitalization of traditional industries are flourishing, with intelligent manufacturing, autonomous driving, intelligent healthcare, and intelligent education having become emerging pillar industries of digital capitalism. At the macro level of the two major departments of production, on the production side, digital capital provides digital-intelligent services for optimizing production processes, predicting demand, and managing supply chains, becoming a new force in the first department of social reproduction; on the consumption side, digital capital provides digital-intelligent services such as online shopping, mobile payments, telemedicine, digital entertainment, and chatbots, becoming a new member of the second department of social reproduction.

Second, the capital circulation of digital capital as a new type of commercial capital.

In circulation-oriented platforms (including intermediary platforms and transaction platforms), digital capital aggregates market participants or creates social connections, opens up digital markets, and forms a digital circulation sphere, greatly assisting traditional producers in the circulation of commodities and the realization of value. From the perspective of the circulation process of industrial capital relying on online sales, digital capital provides intelligent marketing services, performing the intermediary function of commercial capital. According to Marx’s understanding, commercial capital movement is a “special sphere of capital增殖.” Following this understanding, it can be said that, in performing the function of commercial capital, contemporary circulation-oriented platforms have also become a special sphere of capital增殖.

Unlike the traditional understanding described above, the intelligent marketing service provided by digital capitalists is a commodity. Therefore, the peculiarity of digital capital's valorization does not lie, as the "platform rent thesis" holds, in a capital circulation taking the form G—G′ based on the transfer of profits from third parties on the platform. Rather, it lies in the fact that the digital capital controlling the circulation platform possesses the dual attributes of both commercial capital and industrial capital, causing its value movement to assume a "dual circulation" pattern. On the one hand, the digital platform, as the circulation intermediary of digital capitalist social reproduction, promotes and accelerates the value creation and value realization of traditional industrial capital, thereby occupying a subordinate position within the circulation of traditional industrial capital. This is manifested in the purchase phase G—W, where the digital platform aggregates factor resources, opens up online labor markets, and accelerates the preparation of production conditions for traditional industrial capital; and in the sales phase W′—G′, where it provides timely and precise marketing services, accelerating the realization of surplus value for traditional industrial capital. On the other hand, the digital platform also provides various market-oriented digital-intelligence services. The movement of digital capital encompasses both the production process and the circulation process, both value creation and value realization, thereby independently forming a capital circulation as a new type of industrial capital. At this level, the third parties on the platform that produce various commodities for sale become the actual buyers of the digital platform. In fact, in the face of the expanding commercial sector of capitalism and the increasing importance of marketing activities, some research has long pointed out that clarifying the dual nature of commercial capital in simultaneously performing intermediary and production functions, and expanding Marx's concept of productive service commodities, can lead to a better understanding of the real contribution of commercial service activities to the value movement of capitalism. It should be noted that the understanding of Marx's theory of capital circulation needs to be deepened. One cannot remain solely within the sphere of circulation and simply regard the online advertising revenue of digital platforms as a market transfer of surplus value from the real economy sector; otherwise, this would contradict the economic reality of the digitalization, online operation, and intelligentization of productive services in the digital era.

Third, the capital circulation of digital capital in the virtualization of the platform economy.

In the context of political economy, "the formation of fictitious capital is called capitalization." Fundamentally, the capital-like nature of credit money, government bonds, stocks, etc., is based on the fact "that no income has been obtained in advance, but rather on the expectation of a future process of accumulation." The global practice of the platform economy shows that, under the pursuit of the capital market, the actual movement of digital capital from the outset exhibits a significant tendency for "value to detach from the general circulation of capital and circulate in the form of money in the virtual value market." In other words, under the platform economy development model of "growth first, profit later," the circulation of digital capital displays similar characteristics to the movement of fictitious capital. This is prominently manifested in the fact that, since the platform revolution, digital platforms have striven by all means to capture and hoard digital resources, hoping for a future "data gold rush" by promoting the factorization, assetization, and capitalization of data resources, and then leveraging the scale effects and network effects of digital platforms to support the market myth of the exponential expansion of digital capital with the high valuation of digital platforms. This is also manifested in the fact that super digital platforms compete to invest in digital technology and an artificial intelligence arms race, leveraging the capitalization of a new generation of artificial intelligence amid the clamor for the accelerated arrival of general-purpose artificial intelligence, and thereby using carefully woven "future imaginaries" to stir up an artificial intelligence frenzy in the capital market and a digital capital boom in economic development. As a result, under the theoretical protection of technological solutionism and the continuous injection of financial capital (especially venture capital), even platform start-ups with no profitable channels can become the "new darlings" of the capital market, easily realizing the G—G′ capital circulation. It should be noted that the tendency of digital capital circulation to become virtualized, detaching from "production—valorization," is not a frantic pursuit of profit growth without investment, but rather a real reflection of the digital capitalist accumulation system being caught up in data worship and intelligent fetishism. In a certain sense, the disorderly expansion of digital capital and the virtualized development of the platform economy continue the logic of expropriation of financial capitalist accumulation, gestating a digital capitalist economic crisis that will eventually erupt.