Amazon and the New Platform Capitalism
Amazon began as an online bookseller. It has since become something far more consequential: a vast digital infrastructure through which consumers, merchants, advertisers, workers, software developers and businesses increasingly interact. That transformation captures the central idea of platform capitalism. As the original analysis observes, companies such as Amazon, Facebook, Google and Uber have created digitally mediated networks that connect different groups of economic actors and reduce the friction involved in transactions. The resulting model has been associated with the Fourth Industrial Revolution and a new economic order in which data, algorithms, networks and platforms become central sources of value.
The question, however, is no longer simply whether Amazon’s platform model is successful. The more difficult question is whether it is sustainable. The evidence from 2025–26 suggests a paradox. Amazon’s economic model has become substantially stronger and more diversified, yet precisely this expansion has intensified concerns about labour power, advertising practices, data, competition and regulatory control. Amazon’s 2025 results illustrate the scale of the transformation. The company reported $716.9 billion in consolidated net sales, up 12% from 2024. AWS alone generated $128.7 billion, while advertising services generated $68.6 billion and third-party seller services generated $172.2 billion.
Amazon is therefore no longer adequately understood as a retailer. It is simultaneously a marketplace, logistics network, advertising system, cloud-computing infrastructure, subscription ecosystem and increasingly an artificial-intelligence platform. That is the real story of contemporary platform capitalism.
The Platform Does Not Merely Sell Products—It Organises Markets
The original analysis correctly identifies one of the defining characteristics of platforms: they create value by connecting previously separate networks. Amazon connects consumers with merchants; merchants with logistics and fulfilment services; advertisers with consumers; and businesses with computing infrastructure through AWS. The source text describes this as a process in which diverse networks are combined to generate value and reduce information asymmetries. This logic has become even more powerful.
Amazon’s 2025 accounts show that third-party seller services reached $172.2 billion, while AWS reached $128.7 billion. Advertising services reached another $68.6 billion. These businesses are not independent in the conventional sense. They reinforce one another.
More sellers create more product selection. More products attract more consumers. More consumers generate more behavioural and commercial data. More data can improve search, recommendations, logistics and advertising. More advertising revenue can subsidise further investment in the marketplace. And AWS provides the computational infrastructure on which a growing portion of the digital economy operates. This produces a powerful feedback loop. The platform’s competitive advantage therefore comes not simply from selling goods cheaply, but from controlling connections between markets.
This is an important evolution from the traditional industrial corporation. Fordist firms were generally organised around the production of particular products. Platform firms increasingly derive their power from orchestrating ecosystems.
Data Has Become Capital
The original text argues that Amazon possesses two important forms of currency: money and information. User and supplier information becomes an important resource for understanding demand, optimising transactions and strengthening the company’s competitive position.
That argument has become even more relevant in the age of artificial intelligence. AI systems require enormous quantities of data, computing power and distribution infrastructure. Amazon possesses all three at significant scale. The consequence is that the platform economy increasingly resembles a system in which data and infrastructure reinforce one another. The company knows what consumers search for, what products they purchase, how sellers compete for visibility, how goods move through its logistics network and how businesses use its cloud infrastructure. These relationships generate information that can potentially improve the platform’s ability to predict and organise economic activity.
But information asymmetry has a second side. The more information a platform possesses relative to sellers, workers and consumers, the greater the platform’s ability to shape the conditions under which those groups participate. This is where the efficiency argument meets the power argument.
Efficiency and Inequality Can Exist at the Same Time
One of the strongest arguments in favour of platform capitalism is efficiency. Platforms can reduce search costs, transaction costs and coordination costs. A consumer can locate millions of products through a single interface. A small seller can access a global customer base without building its own physical retail infrastructure. A business can rent computing capacity through AWS rather than constructing an equivalent data centre.
The original analysis therefore argues that Amazon can simultaneously weaken traditional bargaining structures while increasing efficiency in product and labour markets. This apparent contradiction is crucial.
Efficiency does not automatically mean equality.
A platform can make a market more efficient while simultaneously increasing the dependence of participants upon the platform. A small merchant may gain access to millions of customers but become increasingly dependent on Amazon’s ranking systems, fees, advertising mechanisms and fulfilment infrastructure.
A worker may gain access to employment through digitally organised work but have less influence over how tasks are allocated, monitored or evaluated. A consumer may receive faster delivery and greater choice while having less visibility into how recommendations, rankings and prices are generated. The platform therefore creates value while also creating dependency.
The Algorithm Becomes the Manager
This issue extends beyond Amazon. Algorithmic management has become a defining feature of modern digital capitalism.
The International Labour Organization defines algorithmic management as the use of algorithms and data-driven systems to organise, allocate, monitor, supervise and evaluate work. It is increasingly used across logistics, transport, customer service and other sectors.
The OECD’s 2025 survey provides an important indication of how rapidly this model is spreading. Among surveyed firms, 90% of US managers reported that their firms had adopted at least one algorithmic-management tool, compared with an average of 79% across France, Germany, Italy and Spain. These systems can allocate work, monitor activity, track time or location, set targets and evaluate performance.
This does not mean that algorithms are inherently harmful.
Managers surveyed by the OECD also reported that algorithmic tools could improve decision-making by increasing the information available to managers, speeding decisions and increasing managerial autonomy. At the same time, they identified concerns involving unclear accountability, lack of explainability and worker health.
The important insight is therefore not that algorithms replace managers. It is that management itself is becoming infrastructural. Workers increasingly interact with systems that determine schedules, allocate tasks, measure performance and sometimes influence access to future work.This gives the platform a form of organisational power that differs from conventional managerial authority.
Amazon’s Labour Question Has Not Disappeared
The original analysis argues that platform capitalism can weaken labour’s bargaining power by presenting workers as independent participants while allowing corporations to exercise significant control through technology and algorithms.
Recent research suggests that this issue remains relevant even as the technology becomes more sophisticated. A 2025 study of Amazon’s workplace practices in Bessemer, Alabama, examined algorithmic management in the context of labour organising and argued that algorithmic systems have become an important terrain of workplace conflict.
More broadly, the ILO’s 2025 work on AI and digitalisation stresses that automation and smart monitoring can improve workplace safety and productivity but can also introduce new risks. It calls for adaptive policies, worker participation and effective risk assessment.
The issue is consequently larger than Amazon. If algorithms increasingly become the interface between workers and employers, the fundamental question becomes: Who controls the rules embedded in the algorithm? And perhaps more importantly: Who can challenge them?
The Advertising Platform Reveals the Hidden Economics of Intermediation
One of the most significant recent developments concerns Amazon’s advertising businesses. In August 2026, the US Federal Trade Commission and 22 states sued Amazon, alleging that the company secretly increased the prices paid by advertisers in its search advertising auctions. The complaint alleges that Amazon represented its Sponsored Products auctions as second-price auctions while introducing an undisclosed surcharge mechanism that caused advertisers to pay substantially more. These are allegations in a pending case, not established findings of fact.
The significance goes beyond the particular legal dispute. It illustrates how platform power can operate through market design. Amazon is not simply participating in an advertising market. It designs the rules through which advertisers compete for visibility on Amazon itself. That creates a structurally unusual position: the platform can simultaneously host the market, collect information about the market, establish the rules of the market and earn revenue from transactions within the market. This is precisely why contemporary platform capitalism cannot be understood simply through the traditional distinction between producer and intermediary. The intermediary can become an economic actor with the capacity to shape the market itself.
Regulation Is Catching Up
The original analysis anticipated that governments would struggle to restrain digital platforms because of their economic importance. Recent developments show both the strength and the limits of that argument. The European Union designated Amazon as one of six digital “gatekeepers” under the Digital Markets Act. The EU has subsequently expanded its scrutiny of digital infrastructure. In June 2026, the European Commission announced a preliminary view that Amazon Web Services should be designated as a DMA gatekeeper, arguing that AWS occupies an important gateway position between businesses and their customers and has an entrenched position supported by factors including switching costs and ecosystem effects. This remains a preliminary position rather than a final determination.
That development is particularly revealing. Regulation is no longer concerned only with the consumer marketplace. It is increasingly concerned with the infrastructure beneath the digital economy. Amazon’s power therefore exists on several layers:
- Marketplace power — connecting sellers and consumers.
- Logistics power — controlling fulfilment and delivery infrastructure.
- Advertising power — determining how sellers purchase visibility.
- Data power — collecting and processing information generated through the ecosystem.
- Cloud power — providing computational infrastructure to other businesses.
- AI power — increasingly connecting computing, models, applications and data.
The platform is therefore becoming an infrastructure stack.
The Sustainability Question Is Now More Complicated
The original argument concludes that Amazon’s ability to integrate platforms and continually enter new markets gives its business model long-term sustainability.
There is considerable evidence supporting the resilience of this model.
Amazon continues to diversify. Its 2025 revenue mix demonstrates that the company is no longer dependent upon one source of income. AWS, advertising, subscriptions and third-party seller services have become substantial businesses alongside retail. Yet sustainability should not be confused with growth.
A business model can be financially durable while generating social, regulatory or environmental pressures that eventually constrain it. Amazon’s own 2025 sustainability reporting illustrates this tension. The company reported that its absolute carbon emissions increased in 2025, even while it said carbon intensity had fallen 38% since 2019 as revenue grew 156%. The distinction matters.
If Amazon continues to grow rapidly, reductions in emissions intensity do not necessarily mean that total environmental pressures decline. The same principle applies to labour and competition. Greater efficiency can coexist with greater dependence. Greater convenience can coexist with greater market concentration. Greater technological sophistication can coexist with greater opacity. Greater scale can coexist with greater regulatory risk.
From Platform Capitalism to Infrastructure Capitalism
The most important insight from the latest evidence may therefore be that the original concept of platform capitalism is itself evolving. The first generation of platforms primarily connected users. The second generation increasingly organises markets. The emerging third generation provides the infrastructure upon which other digital businesses operate. AWS is a particularly important example.
If a retailer, financial company, software developer or AI company relies upon cloud infrastructure, the relationship is no longer simply between a consumer and an online platform. It becomes a relationship between businesses and privately operated digital infrastructure. This is why the European Commission’s attention to cloud computing is significant. Its 2026 preliminary position suggests that policymakers increasingly see cloud services as potential gateways whose economic significance cannot be measured solely through conventional platform metrics. The future of platform capitalism may therefore be less about apps and websites and more about control over the infrastructure of economic coordination.
Conclusion: Sustainable, But Under Conditions
Amazon’s business model has demonstrated extraordinary adaptability. It moved from books to retail, from retail to logistics, from logistics to subscriptions, from subscriptions to advertising, from advertising to cloud computing and increasingly from cloud computing into artificial intelligence. That ability to recombine technologies, markets and networks remains one of its greatest sources of competitive strength.
The original analysis was therefore right to emphasise platform integration and disruptive innovation as central to Amazon’s durability. But the latest evidence adds an important qualification. Amazon’s sustainability cannot be judged solely by its capacity to grow. The more important question is whether its economic power can remain legitimate and contestable as the company becomes increasingly embedded in the infrastructure of everyday economic life. The central tension of platform capitalism is therefore not simply innovation versus regulation, nor efficiency versus labour rights. It is a deeper tension between private coordination and public dependence.
Amazon creates extraordinary value precisely because it coordinates enormous networks. But as those networks become indispensable, the consequences of how Amazon sets prices, governs sellers, manages workers, uses data and deploys algorithms become increasingly important beyond the company itself.
The platform of the future will not merely sell things. It will help determine how markets function. And that is why the ultimate test of Amazon’s business model is not whether it can survive another decade of technological disruption. It is whether an increasingly powerful private infrastructure can continue to generate innovation and efficiency while remaining sufficiently transparent, contestable and accountable to the people and businesses that depend upon it.
Platform capitalism has survived the transition from the industrial economy to the digital economy. Its next challenge is to prove that concentrated digital power can remain sustainable in an economy increasingly built around it.
By Team Cite&Write
References
Amazon, 2025 Form 10-K — revenue, AWS, advertising and third-party seller data. Amazon 2025 Annual Report / 10-K
Federal Trade Commission, 2026 — Amazon advertising case — current allegations concerning Amazon’s advertising-auction practices. FTC: Amazon advertising lawsuit
European Commission, 2026 — AWS and the Digital Markets Act — preliminary position concerning AWS as a potential gatekeeper. European Commission: AWS gatekeeper investigation
OECD, 2025 — Algorithmic Management in the Workplace — evidence on the spread, benefits and risks of algorithmic management. OECD: How widespread is algorithmic management in workplaces?
International Labour Organization, 2025 — AI and Digitalisation at Work — workplace safety, algorithmic management and digitalisation. ILO: Revolutionizing health and safety
International Labour Organization, 2025 — Social Dialogue on AI and Algorithmic Management — evidence on worker participation and governance. ILO: Global case studies of social dialogue on AI and algorithmic management
Amazon Sustainability, 2025 Report — emissions, carbon intensity and sustainability strategy. Amazon Sustainability Reports
Wu, Zhang, Gan & Thomas, 2026 — “Rethinking Platform Capitalism” — recent research on algorithmic labour control in digital platforms.
Wiggin, 2025 — “Weaponizing the Workplace” — research on algorithmic management and labour organising at Amazon