Introduction
A discussion of the crisis of American capitalism should not begin with the claim that the United States economy is on the verge of immediate collapse. Such a claim is neither consistent with economic realities nor necessary for a serious Marxist critique. The United States remains one of the world’s largest and most dynamic economies, a center of global financial capital, and the leading hub of private investment in advanced technologies and artificial intelligence. The U.S. Bureau of Economic Analysis’s final estimate shows that real gross domestic product grew at an annual rate of 2.1 percent in the first quarter of 2026, with investment, exports, government spending, and household consumption all contributing. The issue examined here, therefore, is not crisis as a simple halt in production or an imminent market collapse. It is the structural contradictions of the capitalist mode of production at a stage when the productive forces have become so socialized, knowledge-based, and interdependent that they increasingly conflict with private ownership and private appropriation of the social product.
In this sense, economic growth does not negate crisis; growth itself can carry contradiction. Society produces more wealth, productivity rises, and technology makes it possible to produce with less human labor, while ownership of wealth and capital becomes concentrated in a small part of society and the majority still gain access to the social product mainly by selling their labor power. This article examines the present crisis of American capitalism within that framework. Its point of departure is Marx’s political economy and concepts such as accumulation, concentration of capital, surplus value, financialization, fictitious capital, socialization of production, and the contradiction between productive forces and relations of production. The goal is not to repeat nineteenth-century concepts mechanically, but to examine their renewed meaning under conditions in which knowledge, data, algorithms, networks, and artificial intelligence have become direct productive forces—and to ask whether this transformation may also create the material and cognitive conditions for a different order, referred to here as “a new social order in the age of awareness.”
From Cyclical Crisis to the Structural Contradiction of Capital Accumulation
In mainstream economics, crisis is often treated as a departure from equilibrium. Higher energy prices, supply-chain disruption, unsuitable monetary policy, a pandemic, war, or a financial-market collapse may push the economy off its normal path, and policymakers are expected to restore balance. Marx’s political economy asks a different question: if crises recur throughout capitalist history in many forms, should their source be sought in the logic of accumulation itself? In Marx’s analysis, capital is not merely machinery or money; it is a social relation in which money enters production in order to return as more money. The capitalist purchases labor power and means of production, commodities are produced, and the realization of their value must return the original capital together with surplus value. The ultimate purpose of capitalist production is therefore not the direct satisfaction of social need, but the reproduction of capital on an expanded scale.
Competition compels each capital to cut costs, introduce new technology, raise productivity, and preserve or expand market share. This process creates a contradiction fundamental to understanding capitalist crises. Every firm seeks to reduce labor costs relative to output, yet capital as a whole needs social purchasing power to sell its products. Each firm benefits from replacing human labor with technology, while most people still obtain income by selling their labor power. What is rational from the standpoint of an individual capital can therefore generate imbalance at the level of the system. Capitalism continually expands productive capacity, but no automatic mechanism guarantees that the purchasing power of the majority will rise at the same rate.
This is where overaccumulation becomes important. Accumulated capital must be reinvested at an acceptable rate of profit. When productive investment opportunities do not expand as rapidly as the mass of capital, part of that capital moves into financial markets, real estate, debt securities, and other assets. Credit can temporarily bridge the gap between production and purchasing power, and asset values can rise much faster than real output, but this does not eliminate the contradiction; it postpones it. Marx’s concept of “fictitious capital” is important here: financial claims on future income and value can themselves be bought and sold and appear to grow independently of production, even though their ultimate realization remains dependent on the social production of value.
Financialization, Debt, and the Socialization of Capital Risk
The American economy over recent decades is a leading example of this logic. Financialization does not mean that material production has lost importance; it means that financial relations and asset ownership play a growing role in organizing the economy. A family whose income is insufficient to buy a house takes a mortgage; a student incurs debt for education; households use credit cards for consumption; corporations issue debt to invest, buy back shares, or merge with competitors; and the government borrows to finance expenditures. Credit allows society to replace today’s limitation with a claim on tomorrow’s income, while committing part of that future income in advance to principal and interest payments.
The 2008 financial crisis and the 2020 pandemic exposed the contradictory nature of this structure. When the normal reproduction of capital, credit, or household income was disrupted, the government and central bank mobilized enormous resources to prevent the collapse of payment, credit, and production chains. These experiences weaken the myth of capitalism as a sphere independent of the state. Advanced capitalism is inconceivable without the state’s legal, monetary, financial, scientific, and physical infrastructure. When the survival of the system is threatened, the state becomes an institution for socializing part of the risks created privately. Private ownership confronts the social character of its consequences: the losses of a major financial or productive institution can have such broad effects that society is compelled to bear part of the cost of preserving the conditions for the reproduction of the system as a whole.
Concentration of Wealth and the American Class Structure
To understand who controls the product of this system, analysis must move from aggregate indicators to the structure of ownership. A limitation of public debate about inequality is its frequent focus on income differences. Income and wealth are not the same. Income is a flow received during a period; wealth is a stock of assets that can itself generate income, interest, profit, rent, and capital gains. For class analysis, the distribution of wealth—and especially the kinds of assets owned—is therefore more important than a simple comparison of annual salaries.
Federal Reserve Distributional Financial Accounts for the first quarter of 2026 provide a clear picture of asset concentration. In corporate equities and mutual-fund shares alone, the top 0.1 percent of households owned about $13.33 trillion, the next 0.9 percent about $14.31 trillion, and the next 9 percent about $20.51 trillion. By contrast, the entire bottom 50 percent owned only about $590 billion of these assets. This is not merely a difference in consumption. Stock ownership creates a claim on profit and capital appreciation; concentration of these assets therefore represents concentration of the capacity to appropriate ownership income.
In Marxist analysis, class cannot be reduced to low-, middle-, and high-income groups. A physician, professor, or engineer may earn a high income, but if that income still comes mainly from selling specialized labor power, that person’s relation to production is not the same as that of the owner of a vast concentration of capital. Wealth not only generates income; it reproduces itself. Shareholders receive profit and capital gains, landlords collect rent, bondholders receive interest, and investors can use returns on existing assets to acquire new ones. Households without wealth may instead incur debt for housing, education, or even essential consumption and pay interest rather than receive it. Wealth is thus not only a result of past inequality; it is a mechanism for producing future inequality.
The concentration of wealth must therefore also be understood as a concentration of power. Capital ownership enables its holders to decide where to invest, what technology to use, how many workers to employ, where production will be located, how research budgets are allocated, and in what direction firms develop. Political democracy rests formally on one person, one vote, but economic resources are not allocated on that principle. A person or institution controlling billions of dollars has far greater economic decision-making power than millions of citizens whose principal asset is their labor power. Economic power can also become social and political influence through campaign finance, lobbying, media ownership, support for think tanks and universities, and more direct access to policymakers. This does not mean every political decision is directly dictated by the wealthy; it means that enormous resource inequality also makes the capacity to influence collective decisions profoundly unequal.
From the Concentration of Capital to the Transformation of Productive Forces in the Age of AI
This ownership structure now intersects with a historic transformation of the productive forces. Artificial intelligence, machine learning, data centers, cloud computing, advanced chips, and automation are becoming central components of productive infrastructure. Stanford University’s AI Index 2026 reports that private AI investment in the United States reached $285.9 billion in 2025—more than twenty-three times the $12.4 billion recorded in China. In the same year, 1,953 newly funded AI companies emerged in the United States. These figures should not be treated merely as evidence of a booming new industry. Capital on this scale is moving toward a field that promises greater productivity, lower costs, new markets, and revenue and rent derived from intellectual property and control over infrastructure.
During the Industrial Revolution, investment in mechanical machinery multiplied human physical power; the digital revolution expanded the processing and transmission of information; artificial intelligence now brings parts of human cognitive activity—analysis, translation, programming, design, text and image production, pattern recognition, and decision support—into the sphere of automation. Yet AI differs importantly from the classical industrial machine. Its cognitive raw material is largely humanity’s accumulated social product. Language, books, scientific articles, images, software code, university research, social data, and the knowledge of past generations have created the environment from which AI models draw. Marx’s discussion in the Grundrisse of the growth of “general social knowledge,” later called the General Intellect, therefore acquires renewed significance. AI should not be regarded as a simple mechanical fulfillment of Marx’s prediction, but the structural resemblance is striking: social knowledge can now be fixed in machines, networks, and algorithms and participate directly in production.
At this point, the role of platforms and the conversion of social activity into capital income becomes crucial. Facebook and Instagram within Meta, and YouTube and Google services within Alphabet, are not neutral communication tools; they are privately owned infrastructures that build vast advertising and service markets on user activity, content production, search, viewing, interaction, and social attention. Alphabet reported $294.7 billion in advertising revenue for 2025, including $40.4 billion directly from YouTube advertising; total YouTube revenue from advertising and subscriptions exceeded $60 billion. The company also earns from subscriptions such as YouTube Premium and Google One, which can include access to advanced Gemini models, and from Google Cloud through usage fees and subscriptions for infrastructure, platform, and AI services. Profitability in the digital and AI economy therefore does not arise solely from selling an “AI product.” Advertising, subscriptions, cloud services, access to computing infrastructure, and the conversion of user attention and activity into commercial opportunity form layers of a single system of accumulation.
The class significance of this mechanism lies in the separation between the social activity of millions and billions of people and the concentrated ownership of the infrastructure that organizes and monetizes it. This does not mean that all value generated in the AI economy flows directly to a few platform owners. The value chain includes chips, energy, data centers, networks, cloud computing, foundation models, software, advertising, and consumer services, and capital and profit are distributed across its layers. Nevertheless, firms controlling key bottlenecks—platforms, computing infrastructure, distribution networks, models, or access to users—occupy a privileged position for appropriating part of socially generated value. The increasing socialization of knowledge and data production does not necessarily socialize ownership or its economic product; within existing relations, it can instead create a new form of capital concentration.
The contradiction becomes clearest here. The knowledge used in this process is profoundly social, but the infrastructure that converts it into economic power—data centers, chips, foundation models, cloud networks, and intellectual-property rights—can be privately owned by a small number of companies and investors. Federal Reserve data also show that corporate equity ownership is already concentrated at the top of the wealth distribution. If AI’s economic value is realized mainly through corporate profit and higher share prices, then without a change in ownership relations, much of the benefit of this technological transformation may flow to the same groups that already hold most financial capital.
The classical contradiction between social production and private appropriation thus assumes a new form. Millions—and over generations, billions—of people have contributed to the knowledge and data that now constitute a major raw material of the AI economy. Yet decisions about using this capacity and the appropriation of much of the resulting economic value may remain in the hands of a far smaller group. Under these conditions, artificial intelligence is not merely a technology; it is a question of ownership and therefore a class question.
Automation and the Contradiction Between Liberation from Labor and Economic Insecurity
This transformation produces a deeper contradiction. From a human standpoint, higher productivity and a reduced need for labor should be positive achievements. If society can produce what it needs in half the former working time, the rational outcome could be shorter working hours, more free time, and greater opportunity for education, art, family, social activity, and personal development. Technology has emancipatory potential in this sense. But in a society where most people obtain income mainly by selling their labor power, a reduced need for labor may have an entirely different result. If a company uses AI to produce the same output with fewer workers, higher productivity can mean lower costs and higher profit for the firm, while the same technological advance becomes economic insecurity for a worker who loses a job.
This contradiction is not created by AI. Its source is the social relation that ties access to the social product to wage employment. If automation reaches a point where machines perform a growing share of necessary labor, society confronts a question that is no longer merely technological: if fewer human working hours are needed to produce wealth, why should human security remain dependent on selling ever more hours of labor? Here one may speak of a historical possibility of moving from a “crisis of scarcity” toward a “crisis of potential abundance.” If machines can produce more goods, services, and knowledge with less human labor, the central question is no longer only how to increase production, but who owns productive capacity, who decides its use, and how the gains from productivity are distributed.
Structural Crisis and the Limits of Reform Within Capitalism
This analysis should not imply that economic reforms are unimportant. Progressive taxation, stronger unions, higher wages, public services, universal education and health care, support for the unemployed, and regulation of monopolies can improve millions of lives and alter the balance of power between labor and capital. Marxist critique becomes dogmatic when it dismisses real differences in human living conditions merely because a reform does not immediately abolish capitalism. Yet reforming income distribution is not the same as transforming ownership relations. If the basic structure of capital ownership remains intact, the mechanisms that concentrate wealth can resume operation after a period of reform. Capital reproduces itself through profit, interest, asset appreciation, and intergenerational transfer. The Marxist question is therefore not only how the social product should be distributed more fairly after production, but who owns the means of production, who decides investment, and who participates in setting society’s economic priorities.
This distinction is especially important now because AI can increase the scale of concentration. Advanced models require chips, energy, data centers, networks, and enormous amounts of capital. High fixed costs can make entry difficult and reinforce the advantage of firms that already possess capital, data, and infrastructure. The most socialized form of knowledge production may therefore coexist with one of the most concentrated forms of infrastructure ownership.
From Capitalist Contradiction to the Possibility of a New Social Order in the Age of Awareness
The existence of structural contradiction does not mean that capitalism will collapse automatically. The history of capitalism demonstrates a considerable capacity for adaptation. Keynesian policy, the welfare state, credit, globalization, financialization, monetary policy, privatization, and technology have at different times helped reconstruct the conditions of accumulation. Crisis does not necessarily produce progressive outcomes; it can lead to greater concentration of capital, authoritarianism, nationalism, war, digital surveillance, and new forms of domination. No economic law automatically produces a better social order from capitalism’s intensifying contradictions. This marks an important boundary between historical determinism and the concept of a new social order in the age of awareness. Productive forces can create the possibility of transformation, but possibility is not necessity. Social awareness and human action stand between the two.
Here awareness takes on a meaning beyond individual or moral consciousness. For the first time in history, emerging technologies can potentially process vast quantities of information about production, consumption, resources, needs, and the consequences of economic decisions in near real time. The same AI used to increase private corporate profit could technically be used for resource planning, assessing social needs, making decisions transparent, enabling citizen participation, and coordinating complex production networks. Technology does not determine the social form of its use; relations of power and ownership determine the ends this capacity serves.
From this perspective, a new social order in the age of awareness should not be imagined merely as a new version of nationalization or the transfer of control from private companies to the state apparatus. If economic power passes from a small group of capital owners to a small group of state managers, the problem of human participation in power is not necessarily resolved. The deeper issue is how ownership, decision-making, and the management of social resources can be organized so that people are not merely recipients of economic decisions but conscious participants in shaping them.
This requires rethinking the relationship among ownership, labor, income, and power. If higher productivity makes shorter working time possible, its benefit could be expressed as a reduction in society’s working hours rather than mass unemployment. If knowledge is the product of humanity’s historical cooperation, more social forms of access and ownership can be considered. If technology enables participation and large-scale information processing, structures beyond periodic elections and centralized bureaucratic administration become conceivable. And if production is truly social, participation in economic power is no longer an abstract ideal; it becomes a question of bringing social relations into alignment with the actual character of the productive forces.
Conclusion
The American economy in 2026 presents a contradictory but historically significant picture. The economy continues to grow while ownership of productive assets remains intensely concentrated. In corporate equities and mutual-fund shares, groups at the top of the wealth distribution hold the overwhelming share, while the bottom half of society owns very little. At the same time, with $285.9 billion in private AI investment during 2025, the United States was by far the world’s leading center of private AI investment. These figures do not by themselves “prove” Marx’s theory, and statistics should not be asked to establish more than they can. Taken together, however, they reveal a phenomenon that Marxist analysis still helps explain: production becomes increasingly social, knowledge becomes a direct productive force, and capital depends on ever larger scales of social coordination, while ownership and the right to appropriate the product can remain concentrated in a small part of society.
Artificial intelligence takes this contradiction to a new stage. Humanity’s accumulated social knowledge can become a productive force embodied in models, algorithms, and machines, while the infrastructure that converts that knowledge into economic power remains private. Technology can reduce the need for human labor, but if income remains tied to wage employment, the same advance can create insecurity. Productivity can make greater abundance possible, but if access to the social product remains governed by ownership, potential abundance can coexist with real deprivation.
Under these conditions, capitalist crises may gradually move beyond cyclical market crises and become crises of adaptation between relations of production and the productive forces capitalism itself has created. Yet there is no guarantee that these contradictions will lead to a freer, more just, or more humane order. The same technologies can be used to concentrate power, surveil, and control. The difference between these paths will be determined not by technology alone but by the level of social awareness, organization, democratic institutions, and people’s capacity to participate consciously in shaping their future.
The possibility of a new social order in the age of awareness emerges from this point—not as an inevitable stage toward which history will move on its own, but as a possibility whose material conditions are created by the development of the productive forces and whose realization depends on conscious social action. The difference between this view and historical determinism lies precisely here: the new order will not rise automatically from capitalism’s ruins; it must be consciously conceived, designed, and built through social participation.
In such a horizon, the economy is no longer society’s end but a means to human flourishing; technology is not a means of removing people to raise profit, but a means of reducing unnecessary labor and expanding freedom; knowledge is not merely a commodity to be monopolized, but a social inheritance; and economic power is not a natural right derived from the amount of property one owns, but a sphere in which society must be able to participate consciously. By socializing production, globalizing the division of labor, transforming knowledge into a productive force, and now reshaping the productive forces through AI, capitalism itself has created the material conditions for a question of growing importance in the twenty-first century: if knowledge, production, and wealth are more than ever the products of human social activity, why should ownership and decision-making power over them remain so private and concentrated?
The answer is not predetermined. Capitalist crisis does not guarantee the next order. But as the distance widens between the social character of the productive forces and the private concentration of ownership, the possibility of a historical alternative becomes more real. “A new social order in the age of awareness” is one name for this possibility: a society that, instead of waiting for history to impose an order upon it, consciously decides what order it wants to build.
Principal Sources
- Marx, Karl. Capital: A Critique of Political Economy, Vols. I–III.
- Marx, Karl. Grundrisse: Foundations of the Critique of Political Economy, especially “Fragment on Machines.”
- Federal Reserve Board. Distributional Financial Accounts, 2026 Q1.
- U.S. Bureau of Economic Analysis. GDP (Third Estimate), Industries, Corporate Profits, State GDP, and State Personal Income, First Quarter 2026.
- Stanford Institute for Human-Centered Artificial Intelligence. Artificial Intelligence Index Report 2026.
- Alphabet Inc. Annual Report / Form 10-K for fiscal year 2025.
