Is the AI bubble hiding capitalism’s deeper crisis?

The surge in artificial intelligence (AI) investment reflects monopoly-finance capitalism adapting to deeper economic strain, rather than the arrival of a techno-feudal order, according to a new analysis by Elliot Goodell Ugalde of Queen's University. The study asserts that the AI boom is being driven less by proven productivity gains than by a deeper need within mature capitalism to absorb excess capital, sustain speculative expectations and reorganize power around a small group of dominant technology firms.

The article, titled Beyond techno-feudalism: A monopoly-capital analysis of artificial intelligence investment, was published in Capital & Class. It challenges popular claims that platform power, digital rent extraction and data enclosure mark a shift beyond capitalism, arguing instead that AI investment reflects long-running capitalist dynamics of overaccumulation, financial expansion, monopoly concentration and profitability strain.

AI investment is framed as surplus absorption, not techno-feudal rupture

Techno-feudalism, a term associated with claims that digital platforms operate like modern fiefdoms, argues that market competition is being displaced by platform control, rent extraction and dependence on privately owned digital infrastructures. Under this view, cloud systems, algorithms, app stores, data architectures and platform ecosystems give large technology companies a kind of command power that resembles feudal rule more than capitalist exchange.

The author challenges this argument. The analysis does not deny the power of platforms, the rise of digital rents or the growing dependence of firms and users on proprietary infrastructures. Rather, it argues that these developments remain internal to capitalism. Platform rents, cloud fees, data access costs and licensing charges are presented as advanced forms of capitalist monopoly power, not evidence of a postcapitalist order.

The article shifts the focus from the visible power of platforms to the deeper economic conditions that make AI such an attractive destination for capital. The key claim is that AI investment has become a mechanism for absorbing surplus capital at a time when mature capitalist economies face chronic overaccumulation and limited profitable outlets for productive investment.

In this framework, AI investment is not limited to venture capital funding for AI start-ups. The study defines it more broadly as a composite process involving private equity and venture capital flows, public stock market capitalization of AI-intensive companies, massive fixed capital spending on data centers and semiconductors, energy infrastructure expansion, and state procurement and industrial policy directed toward AI development. Taken together, these flows form a large investment circuit that can absorb capital even when the productive returns remain uncertain.

The analysis is based on the monopoly-capital framework developed by Paul Baran and Paul Sweezy. In that tradition, the key problem of advanced capitalism is not simply producing surplus, but finding ways to absorb it. When profitable opportunities in productive accumulation are constrained, capital seeks outlets in finance, speculation, military spending, advertising, infrastructure and concentrated growth sectors. The study argues that AI has become one such outlet.

On one side, AI is a real technology with potential applications in production, logistics, software, military systems, marketing, administration and data processing. On the other, it is also a speculative investment object whose valuations often depend on future expectations rather than present productivity gains. The article argues that this dual role is central to understanding the boom.

The study challenges both techno-optimist and techno-feudalist narratives. Techno-optimists see AI as a general-purpose technology that will create a new productivity regime. Techno-feudalists see AI and platform power as evidence that capitalism has been superseded. The author argues that both accounts miss the underlying pattern: AI is functioning as a pressure valve for monopoly-finance capitalism, allowing capital to defer crisis by channeling investment into a narrow but highly valued technological enclave.

This also changes the political stakes. If AI represents a new techno-feudal order, the problem is the replacement of markets by digital fiefdoms. But if AI reflects monopoly-finance capitalism, the problem is capital itself: concentrated ownership, speculative valuation, labor displacement, state-backed investment and the repeated transfer of crisis costs onto workers.

The AI bubble rests on speculation, state support and concentrated growth

The study identifies three mechanisms through which AI absorbs surplus capital.

Class-mediated consumption and the expanded sales effort

In traditional monopoly-capital theory, the sales effort refers to advertising, marketing and demand creation. In the AI era, the article argues, this logic extends into digital transformation programs, consulting services, integration projects, corporate AI pilots, procurement strategies and institutional spending meant to signal technological relevance.

Many organizations adopt AI not because it has clearly raised productivity, but because not adopting it may appear risky or backward. Firms, universities, hospitals, government agencies and financial institutions face pressure to show that they are participating in the AI transition. This creates demand for AI tools, software subscriptions, cloud services, consultants, compliance systems and internal restructuring, even when the practical benefits remain uncertain.

The result is a form of investment that can generate revenues and contracts without necessarily delivering proportional productivity gains. AI adoption becomes part of the economy's sales effort, converting expectations into spending and keeping capital in motion. The study argues that this demand is economically functional even when it falls short of its advertised goals.

Investment driven by accumulation itself rather than demonstrated productive need

In this account, AI investment is not primarily a response to proven demand for useful AI outputs. It is also driven by the need to find a place for capital to go. Investors, corporations and states pour money into AI because the sector promises future growth, market dominance and strategic advantage.

This is where the analysis connects AI to fictitious capital. AI valuations often capitalize expected future productivity, even when current returns are limited or uneven. Companies can attract investment based on projected dominance, anticipated automation, future enterprise adoption or strategic infrastructure control. The article argues that this converts future expectations into present financial value.

State support strengthens that pattern. The study points to public procurement, subsidies and strategic industrial policy as central to the AI investment boom. AI is increasingly framed not only as a commercial technology, but as a national security priority. That framing makes it easier to justify public funding, defense contracts, cloud infrastructure procurement, semiconductor subsidies and energy-system expansion tied to AI.

The study argues that this state-backed dimension echoes older monopoly-capital patterns, especially the role of military expenditure as a stabilizing outlet for surplus capital. AI has become part of a new military-industrial and technological complex, where defense demand, cloud procurement and geopolitical competition help sustain investment even when commercial profitability remains uncertain.

Disproportionate sectoral expansion

AI has become a narrow technological enclave that attracts unusually large amounts of capital relative to the broader economy. Growth becomes concentrated in AI-linked companies, semiconductor firms, cloud providers, data center construction, power infrastructure and related consulting or compliance services. This can create the appearance of broad economic momentum while leaving underlying productivity growth weak.

The study argues that such enclave-led expansion is not new. Capitalist economies have repeatedly channeled surplus into leading sectors during periods of stagnation or profitability pressure. What is specific today is the AI-centered nature of the enclave and the scale of capital tied to computing infrastructure, data centers, energy systems and financial expectations around automation.

This concentration also deepens monopoly power. A small number of hyperscale firms control cloud infrastructure, data-bearing environments, compute capacity, platform distribution and AI deployment channels. Smaller AI firms often depend on those giants for computing resources, data access or market reach. The dependence may look like digital vassalage to techno-feudalist theorists, but the study argues it is better understood as monopoly capitalism: large firms using market power, infrastructure control and capital intensity to shape the terms of accumulation.

Workers may bear the cost if the AI investment cycle turns

What may happen when the AI boom reaches its limits? The author claims that the likely resolution is not a postcapitalist transformation, but a familiar capitalist pattern: devaluation, consolidation and intensified class power. The reason lies in AI's role as both surplus sink and labor-displacing technology.

AI requires large investments in constant capital, including data centers, chips, cloud systems, energy infrastructure and software platforms. It is also designed to reduce or reorganize labor in administrative, cognitive, technical and professional work. This combination raises the capital intensity of production while placing downward pressure on employment, wages and bargaining power.

In Marxist terms, the study frames this as an increase in the organic composition of capital. More capital is invested in machinery, infrastructure and technology relative to living labor, the source of surplus value in Marx's theory. Even if individual firms gain advantages through automation, the broader spread of labor-displacing technology can intensify profitability pressures across the system.

The study also highlights the risk of moral depreciation, where fixed capital loses value before it is physically worn out because newer technology makes it obsolete. In the AI sector, chips, data center systems, model architectures and training infrastructure can become outdated quickly. That creates pressure for constant reinvestment just to avoid falling behind. The boom can therefore absorb huge amounts of capital, but it also builds in the conditions for future devaluation.

If expected returns fail to materialize, the correction could be severe. Companies may write down assets, cut staff, cancel projects and consolidate around the largest firms. Smaller AI companies may collapse or be acquired. Workers may face layoffs, wage pressure and intensified workplace monitoring. Public budgets could also be affected if states have heavily subsidized AI infrastructure or if financial instability reduces tax revenues.

The study argues that this pattern would not mark the failure of AI as a technology. AI may continue to develop and remain widely used after a bubble correction, just as the internet survived the dot-com crash. The issue is not whether the technology disappears. It is who pays for the speculative cycle and how ownership becomes reorganized after the crash.

The article suggests that capital destruction is rarely borne equally. Investors and dominant firms may emerge stronger through consolidation, while workers and public institutions absorb much of the disruption. The AI boom could therefore leave behind a more concentrated, more capital-intensive and more labor-displacing economy, even if many speculative firms fail.

The study argues that the scale of AI investment should not be confused with proof of durable productivity growth. High spending on data centers, chips and AI software may raise measured investment and market valuations, but it does not automatically translate into broad-based economic gains.

It also challenges the idea that platform rent extraction means capitalism has ended. For the author, rent, enclosure and monopoly power are not outside capitalism. They are part of how contemporary capitalism manages crisis. AI does not represent a clean break from capitalism's logic; it reflects that logic under conditions of financialization, overaccumulation and technological concentration.

The study does not deny that AI can be productive or that it can reshape industries. However, it insists that AI's economic role must be understood through the system that funds, owns and deploys it. A technology developed under monopoly-finance capitalism will tend to reflect the priorities of monopoly-finance capitalism: capital absorption, market control, labor discipline, state-backed growth and speculative valuation.